Soln Manual 5e

Copyright © 2009 John Wiley & Sons, Inc. All Rights Reserved.

Solutions Manual for: Communications Systems, 5th edition by Karl Wiklund, McMaster University, Hamilton, Canada Michael Moher, Space-Time DSP Ottawa, Canada and Simon Haykin, McMaster University, Hamilton, Canada Published by Wiley, 2009.


Chapter 2

2.1 (a)

( ) cos(2 ) ,2 2

1

c

c

T Tg t A f t t

fT

π −⎡ ⎤= ∈⎢ ⎥⎣ ⎦

=

We can rewrite the half-cosine as:

cos(2 ) rectctA f tT

π ⎛ ⎞⋅ ⎜ ⎟⎝ ⎠

Using the property of multiplication in the time-domain:

[ ]1 2( ) ( ) ( )

1 sin( ) ( ) ( )2 c c

G f G f G ffTf f f f AT

fTπδ δ

π

= ∗

= − + + ∗

Writing out the convolution:

[ ]sin( )( ) ( ( ) ( ( )2

sin( ( ) ) sin( ( ) ) 1 = 2 2

cos( ) cos( ) 1 122 2

c c

c cc

c c

AT TG f f f f f dT

f f T f f TA ff f f f T

A fT fT

f fT T

πλ δ λ δ λ λπλ

π ππ

π ππ

∞

−∞

⎛ ⎞= − + + − −⎜ ⎟⎝ ⎠

⎛ ⎞+ −= +⎜ ⎟+ −⎝ ⎠

⎛ ⎞⎜ ⎟

= −⎜ ⎟⎜ ⎟− +⎝ ⎠

∫

(b)By using the time-shifting property:

0 0 0( ) exp( 2 ) 2

cos( ) cos( )( ) exp( )1 122 2

Tg t t j ft t

A fT fTG f j fTf f

T T

π

π π ππ

− − =

⎛ ⎞⎜ ⎟

= − ⋅ −⎜ ⎟⎜ ⎟− +⎝ ⎠


(c)The half-sine pulse is identical to the half-cosine pulse except for the centre frequency and time-shift.

12cf Ta

=

cos( ) cos( )( ) (cos( ) sin( ))

2

cos(2 ) cos(2 ) sin(2 ) sin(2 ) 4

exp( 2 ) exp( 2 ) 4

c c

c c c c

c c

A fTa fTaG f fTa j fTaf f f f

A fTa fTa fTa fTaj jf f f f f f f f

A j fTa j fTaf f f f

π π π ππ

π π π ππ

π ππ

⎡ ⎤= − ⋅ −⎢ ⎥− +⎣ ⎦

⎡ ⎤= − + −⎢ ⎥− + − +⎣ ⎦

⎡ ⎤− −= −⎢ ⎥− +⎣ ⎦

(d) The spectrum is the same as for (b) except shifted backwards in time and multiplied by -1.

cos( ) cos( )( ) exp( )1 122 2

exp( 2 ) exp( 2 ) 1 142 2

A fT fTG f j fTf f

T T

A j fT j fT

f fT T

π π ππ

π ππ

⎛ ⎞⎜ ⎟

= − ⋅⎜ ⎟⎜ ⎟− +⎝ ⎠⎡ ⎤⎢ ⎥

= −⎢ ⎥⎢ ⎥− +⎣ ⎦

(e) Because the Fourier transform is a linear operation, this is simply the summation of the results from (b) and (d)

exp( 2 ) exp( 2 ) exp( 2 ) ( 2 )( ) 1 142 2

cos(2 ) cos(2 ) 1 122 2

A j fT j fT j fT j fTG ff f

T T

A fT fT

f fT T

π π π ππ

π ππ

⎡ ⎤⎢ ⎥+ − + −

= −⎢ ⎥⎢ ⎥− +⎣ ⎦⎡ ⎤⎢ ⎥

= −⎢ ⎥⎢ ⎥− +⎣ ⎦


2.2

( )( )

( )

( ) exp( )sin(2 )u( ) exp( )u( ) sin(2 )

1 1( ) ( ) ( )1 2 2

1 1 1 2 1 2 ( ) 1 2 ( )

c

c

c c

c c

g t t f t tt t f t

G f f f f fj f j

j j f f j f f

ππ

δ δπ

π π

= −

= −

⎡ ⎤∴ = ∗ − − +⎢ ⎥+ ⎣ ⎦

⎡ ⎤= −⎢ ⎥+ − + +⎣ ⎦

2.3 (a)

[ ]

[ ]

( ) ( ) ( )1( ) ( ) ( )2

( ) rect2

1( ) ( ) ( )2

1 12 2( ) rect rect

e o

e

e

o

o

g t g t g t

g t g t g t

tg t AT

g t g t g t

t T t Tg t A

T T

= +

= + −

⎛ ⎞= ⎜ ⎟⎝ ⎠

= − −

⎛ ⎞⎛ ⎞ ⎛ ⎞− +⎜ ⎟⎜ ⎟ ⎜ ⎟= −⎜ ⎟⎜ ⎟ ⎜ ⎟

⎜ ⎟⎜ ⎟ ⎜ ⎟⎜ ⎟⎝ ⎠ ⎝ ⎠⎝ ⎠


(b) By the time-scaling property g(-t) G(-f)

[ ]

[ ]

[ ]

[ ]

1( ) ( ) ( )21 sinc( ) exp( 2 ) sinc( ) exp( 2 )2

sinc( )cos( )

1( ) ( ) ( )21 sinc( ) exp( 2 ) sinc( ) exp( 2 )2

sinc( )sin( )

e

o

G f G f G f

fT j fT fT j fT

fT fT

G f G f G f

fT j fT fT j fT

j fT fT

π π

π

π π

π

= + −

= − +

=

= − −

= − −

= −


2.4. We need to find a function with the stated properties. We can verify that:

( ) sgn( ) u( ) u( )G f j f j f W j f W= − + − − − − meets the stated criteria. By duality g(f) G(-t)

1 1 1 1 1( ) ( ) exp( 2 ) ( ) exp( 2 )2 2 2 2

1 sin(2 ) 2

g t j t j Wt j t j Wtt j t j t

Wtjt t

δ π δ ππ π π

ππ π

⎛ ⎞ ⎛ ⎞= + − − − −⎜ ⎟ ⎜ ⎟

⎝ ⎠ ⎝ ⎠

= +

2.5 By the differentiation property:

[ ]

( ) 2 ( )

1 ( ) exp( 2 ) ( )exp( 2 )

2 ( )sin(2 )

dg tF j fG fdt

H f j f H f j f

j H f f

π

π τ π ττ

π ττ

⎛ ⎞ =⎜ ⎟⎝ ⎠

= − −

=

But 2 2( ) exp( )H f fτ π τ= −

2 2

2 2

2 2

0

1( ) exp( )sin(2 )

sin(2 ) exp( )

2 exp( )sinc(2 )

lim ( ) 2 sinc(2 )

G f f fTf

fTff

T f fT

G f T fTτ

π τ ππ

ππ τπ

π τ π

π→

∴ = −

= −

= −

=

2

2

0

0

1( ) exp

1 1 ( ) ( )

( ) 1 1( ) ( )

t T

t T

t T

t T

ug t du

h d h d

dg t h t T h t Tdt

πτ τ

τ τ τ ττ τ

τ τ

+

−

+

−

⎛ ⎞= −⎜ ⎟

⎝ ⎠

= +

= − − + +

∫

∫ ∫


2.6 (a) If g(t) is even and real then

* * *

* *

*

1( ) [ ( ) ( )]2

1 1( ) ( )2 2

( ) ( )( ) is all real

G f G f G f

G f G f

G f G fG f

= + −

= −

=∴

If g(t) is odd and real then

* * *

* *

*

1( ) [ ( ) ( )]21 1( ) ( ) ( )2 2

( ) ( )( ) ( )( ) must be all imaginary

G f G f G f

G f G f G f

G f G fG f G f

G f

= − −

= − −

= − −

= −∴

(b)

The previous step can be repeated n times so:

( )

( )

( 2 ) ( ) ( )

But each factor ( 2 ) represents another differentiation.

( ) ( )2

Replacing with

( ) ( )2

nn

n

nn n

nn n

dj ft G t g fdfj ft

jt G t g f

g h

jt h t H f

π

π

π

π

− −

−

⎛ ⎞⋅ −⎜ ⎟⎝ ⎠

⎛ ⎞⎜ ⎟⎝ ⎠

[ ]* *

1( ) ( ) ( )2

and ( ) ( ) ( ) ( )

g t g t g t

g t g t G f G f

= + −

= ⇒ = −

[ ]* *

1( ) ( ) ( )2

and ( ) ( ) ( ) ( )

g t g t g t

g t g t G f G f

= − −

= ⇒ = −

( 2 ) ( ) ( ) by duality

( ) ( )2

dj t G t g fdf

j dt G t g fdf

π

π

− −

⋅ −


(c)

Let ( )( ) ( ) and ( ) ( )2

nn njh t t g t H f G f

π⎛ ⎞= = ⎜ ⎟⎝ ⎠

( )( ) (0) (0)2

nnjh t dt H G

π

∞

−∞

⎛ ⎞= = ⎜ ⎟⎝ ⎠∫

(d)

1 1*2 2

( ) ( )

( ) ( )

g t G f

g t G f−

1 2 1 2

*1 2 1 2

1 2

( ) ( ) ( ) ( )

( ) ( ) ( ) ( ( ))

( ) ( )

g t g t G G f d

g t g t G G f d

G G f d

λ λ λ

λ λ λ

λ λ λ

∞

−∞

∞

−∞

∞

−∞

−

− −

= −

∫

∫

∫

(e)

*1 2 1 2

*1 2

*1 2 1 2

*1 2 1 2

( ) ( ) ( ) ( )

( ) ( ) (0)

( ) ( ) ( ) ( 0)

( ) ( ) ( ) ( )

g t g t G G f d

g t g t dt G

g t g t dt G G d

g t g t dt G G d

λ λ λ

λ λ λ

λ λ λ

∞

−∞

∞

−∞

∞ ∞

−∞ −∞

∞ ∞

−∞ −∞

−

−

∫

∫

∫ ∫

∫ ∫


2.7 (a) 2

2

( ) sinc ( )

( )

max ( ) (0) sinc (0)

The first bound holds true.

g t AT fT

g t dt AT

G f GATAT

∞

−∞

=

=

==

∴

∫

(b)

2

( ) 2

2 ( ) 2 sinc ( )

sin( ) sin( ) 2

sin( ) 2 sin( )

dg t dt Adt

j fG f fAT fT

fT fTfATfT fT

fTA fTfT

π π

π πππ π

π ππ

∞

−∞

=

=

= ⋅

= ⋅

∫

But,

sin( ) 1 and sinc( ) 1

sin( )2 sin( ) 2

2 ( ) 2

fT f fT f

fTA fT AfT

j fG f A

π π

π ππ

π

≤ ∀ ≤ ∀

∴ ⋅ ≤

∴ ≤


2.7 c)

2 2 2

22 2

2

2

( 2 ) ( ) 4 ( )

sin ( ) 4( )

4 sin ( )

4

j f G f f G f

fTf ATfT

A fTTA

T

π π

πππ

π

=

=

=

≤

The second derivative of the triangular pulse is plotted as:

Integrating the absolute value of the delta functions gives:

2

2

22

2

( ) 4

( )( 2 ) ( )

d g t Adtdt T

d g tj f G f dtdt

π

∞

−∞

∞

−∞

=

∴ ≤

∫

∫


2.8. (a)

1 2 1 2

2 1

( ) ( ) ( ) ( ) ( ) ( ) by the commutative property of multiplicationg t g t G f G f

G f G f∗

=

b)

[ ] [ ]

[ ] [ ][ ] [ ]

1 2 3 1 2 3

1 2 3 1 2 3

1 2 3 1 2 3

( ) ( ) ( ) ( ) ( ) ( )Because multiplication is commutative, the order of the multiplicationdoesn't matter.

( ) ( ) ( ) ( ) ( ) ( )

( ) ( ) ( ) ( ) ( ) ( )

g f g f g f G f G f G f

G f G f G f G f G f G f

G f G f G f g f g f g f

∗ ∗

∴ =

∴ ∗ ∗

c) Taking the Fourier transform gives:

[ ]1 2 3

1 2 2 3 1 2 1 2

( ) ( ) ( )Multiplication is distributive so:

( ) ( ) ( ) ( ) ( ) ( ) ( ) ( )

G f G f G f

G f G f G f G f g t g t g t g t

+

+ +


2.9 a) Let 1 2( ) ( ) ( )h t g t g t= ∗

( )

( )

[ ]

1 2

1 2

11 2 2

11 2 2

( ) 2 ( )

2 ( ) ( ) 2 ( ) ( )

( )2 ( ) ( ) ( )

( )( ) ( ) ( )

dh t j fH fdt

j fG f G fj fG f G f

dg tj fG f G f g tdt

dg td g t g t g tdt dt

π

ππ

π

=

=

⎡ ⎤ ∗⎢ ⎥⎣ ⎦⎡ ⎤∴ ∗ = ∗⎢ ⎥⎣ ⎦

b) 2.10.

1 21 2 1 2

11 2 2

11 2

1 2 1

(0) (0)1( ) ( ) ( ) ( ) ( )2 2

(0)1 ( ) ( ) ( ) ( )2 2

(0)1 ( ) ( ) ( )2 2

( ) ( ) ( )

t G Gg t g t dt G f G f fj f

GG f G f f G fj f

GG f f G fj f

g t g t dt g t

δπ

δπ

δπ

−∞

−∞

∗ +

⎡ ⎤ ⎡ ⎤= +⎢ ⎥ ⎢ ⎥⎣ ⎦⎣ ⎦⎡ ⎤

= +⎢ ⎥⎣ ⎦

∴ ∗ =

∫

2 ( )t t

g t−∞

⎡ ⎤∗⎢ ⎥

⎣ ⎦∫ ∫

( ) ( ) ( )t

Y f X X f dν ν ν−∞

= −∫

( )( )( )( )

[ ]

( ) 0 if

( ) 0 if

for when 0 and

for when 0 and

for 0 when 2

for - 0 when 2

Over the range of integration , , the integr

X W

X f f W

f W f W W

f W f W W

f W W f W

f W W f W

W W

ν ν

ν ν

ν ν ν ν

ν ν ν ν

ν ν

ν ν

≠ ≤

− ≠ − ≤

− ≤ ≤ + ≥ ≤

− ≥ − ≤ − + ≤ ≥ −

∴ − ≤ ≤ ≤ ≤

− ≥ − ≤ ≤ ≥ −

∴ − al is non-zero if 2f W≤


2.11 a) Given a rectangular function: 1( ) rect tg tT T

⎛ ⎞= ⎜ ⎟⎝ ⎠

, for which the area under g(t) is

always equal to 1, and the height is 1/T. 1 rect sinc( )t fTT T

⎛ ⎞⎜ ⎟⎝ ⎠

Taking the limits:

0

0

1lim rect ( )

1lim sinc( ) 1

T

T

t tT T

fTT

δ→

→

⎛ ⎞ =⎜ ⎟⎝ ⎠

=

b) 2.12.

1 1( ) sgn( )2 2

By duality:1 1( ) ( )2 2

1( ) ( )2 2

G f f

G f tj t

jg t tt

δπ

δπ

= +

− −

∴ = +

( ) 2 sinc(2 )

2 sinc(2 ) rect2

g t W WtfW WtW

=

⎛ ⎞⎜ ⎟⎝ ⎠

lim 2 sinc(2 ) ( )

2lim rect 12

W

W

W Wt t

W

δ→∞

→∞

=

⎛ ⎞ =⎜ ⎟⎝ ⎠


2.13. a) By the differentiation property: ( )2

2 2

2 ( ) exp( 2 )

1( ) exp( 2 )4

i ii

i ii

j f G f k j ft

G f k j ftf

π π

ππ

= −

∴ = − −

∑

∑

b)the slope of each non-flat segment is:b a

At t

±−

[ ]

( ) [ ]

2 2

2 2

1( ) exp( 2 ) exp( 2 ) exp( 2 ) exp( 2 )4

cos(2 ) cos(2 )2

b a a bb a

b ab a

AG f j ft j ft j ft j ftf t tA ft ft

f t t

π π π ππ

π ππ

⎛ ⎞⎛ ⎞= − − − +⎜ ⎟⎜ ⎟ −⎝ ⎠⎝ ⎠

= − −−

But: [ ]1sin( ( ))sin( ( )) cos(2 ) cos(2 )2b a b a a bf t t f t t ft ftπ π π π− + = − by a trig identity.

[ ]2 2( ) sin( ( ))sin( ( ))( ) b a b a

b a

AG f f t t f t tf t t

π ππ

∴ = − +−

2.14 a) let g(t) be the half cosine pulse of Fig. P2.1a, and let g(t-t0) be its time-shifted counterpart in Fig.2.1b

( )( )( )( )

*

2

2*0 0 0 0

2*0 0

( ) ( )

( )

( ) exp( 2 ) ( ) exp( 2 ) ( ) exp( 2 )exp( 2 )

( ) exp( 2 ) ( ) exp( 2 ) ( )

G f G f

G f

G f j ft G f j ft G f j ft j ft

G f j ft G f j ft G f

ε

π π π π

π π

=

=

− = −

− =


2.14 b)Given that the two energy densities are equal, we only need to prove the result for one. From before, it was shown that the Fourier transform of the half-cosine pulse was:

[ ] 1sinc(( ) ) sinc(( ) ) for 2 2c c c

AT f f T f f T fT

+ + − =

After squaring, this becomes:

2 22 2

2 2 2 2

sin ( ( ) ) sin ( ( ) ) sin( ( ) )sin( ( ) )24 ( ( ) ) ( ( ) ) ( )( )

c c c c

c c c c

f f T f f T f f T f f TA Tf f T f f T T f f f fπ π π π

π π π⎡ ⎤+ − + −

+ +⎢ ⎥+ − + −⎣ ⎦

The first term reduces to:

( ) ( )( )

22 2

2 2 22 2

sin cos cos2

2 2c

fT fT fTT f ffT fT

ππ π π

ππ ππ π

⎛ ⎞+⎜ ⎟⎝ ⎠ = =

+⎛ ⎞ ⎛ ⎞+ +⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠

The second term reduces to:

( )( )

22

2 22 2

sin cos2

2c

fT fTT f ffT

ππ π

πππ

⎛ ⎞−⎜ ⎟⎝ ⎠ =

−⎛ ⎞−⎜ ⎟⎝ ⎠

The third term reduces to:

2

2 22 2 2

2

2 2 22

sin( ( ) )sin( ( ) ) cos( ) cos (2 )21( )( )

41 cos(2 )

14

c c

c c

f f T f f T fTT f f f f T f

TfT

T fT

π π π ππ π

π

π

+ − −=

+ − ⎛ ⎞−⎜ ⎟⎝ ⎠

− −=

⎛ ⎞−⎜ ⎟⎝ ⎠

2

2 2 22

2cos ( ) 1

4

fT

T fT

π

π= −

⎛ ⎞−⎜ ⎟⎝ ⎠

Summing these terms gives:

( ) ( )2 22 2 2

2 22 2

cos cos cos ( )21 14 1 1

2 22 2

fT fTA T fTT f ff f T TT T

π π ππ

⎡ ⎤⎢ ⎥⎢ ⎥+ −⎢ ⎥⎛ ⎞⎛ ⎞⎛ ⎞ ⎛ ⎞ + −+ − ⎜ ⎟⎜ ⎟⎢ ⎥⎜ ⎟ ⎜ ⎟ ⎝ ⎠⎝ ⎠⎝ ⎠ ⎝ ⎠⎣ ⎦


2.14 b)Cont’d By rearranging the previous expression, and summing over a common denominator, we get:

( )

( )

2 2 2

22 22

2

2 2 2

2 4 22 24

2 2 2

22 2 2

cos ( )4 1

4

cos ( )1 14 4 1

16

cos ( )

4 1

A T fTT

fT

A T fTT T f

T

A T fT

T f

ππ

ππ

ππ

⎡ ⎤⎢ ⎥⎢ ⎥⎢ ⎥⎛ ⎞−⎢ ⎥⎜ ⎟⎝ ⎠⎣ ⎦⎡ ⎤⎢ ⎥

= ⎢ ⎥⎢ ⎥−⎣ ⎦⎡ ⎤⎢ ⎥=⎢ ⎥−⎣ ⎦


2.15 a)The Fourier transform of ( ) 2 ( )dg t j fG fdt

π

Let ( )'( ) dg tg tdt

=

By Rayleigh’s theorem: 2 2( ) ( )g t dt G f df∞ ∞

−∞ −∞

=∫ ∫

( )

( )

( )( )

( )

2 22 22 2

22

22 *

222

22 * *

222

2*

22 *

( ) ( )

( )

( ) '( ) ' ( )

4 ( )

( ) '( ) ( ) ' ( )

16 ( )

( ) ( )

16 ( ) ( )

t g t dt f G f dfW T

g t dt

t g t dt g t g t dt

g t dt

t g t g t tg t g t dt

g t dt

dt g t g t dtdt

g t g t dt

π

π

π

⋅∴ =

⋅=

⎡ ⎤−⎣ ⎦≥

⎡ ⎤⋅⎢ ⎥⎣ ⎦=

∫ ∫∫

∫ ∫∫

∫∫

∫

∫

Using integration by parts, we can show that:

2 2

2 22

( ) ( )

116

14

dt g t dt g tdt

W T

WT

π

π

∞ ∞

−∞ −∞

⋅ =

∴ ≥

∴ ≥

∫ ∫


2.15 b) For 2( ) exp( )g t tπ= − 2

2 2 2 2

2 2

2

( ) exp( )

exp( 2 ) exp( 2 )

exp( 2 )

g t f

t t dt f f dfW T

t dt

π

π π

π

∞ ∞

−∞ −∞∞

−∞

−

− ⋅ −∴ =

−

∫ ∫

∫

Using a table of integrals:

2 2

0

2 2

2 2

2

2

2 2

2

1exp( ) for 04

1 1exp( 2 )4 2

1 1 exp( 2 )4 2

1 exp( 2 )2

1 14 2

12

1 4

14

x ax dx aa a

t t dt

f t df

t

T W

TW

π

ππ

ππ

π

π

π

π

∞

∞

−∞

∞

−∞

∞

−∞

− = >

∴ − =

− =

− =

⎛ ⎞⎜ ⎟⎝ ⎠∴ =

⎛ ⎞= ⎜ ⎟⎝ ⎠

∴ =

∫

∫

∫

∫


2.16.

Given: 2( ) and ( ) , which implies that ( )x t dt h t dt h t dt∞ ∞ ∞

−∞ −∞ −∞

< ∞ < ∞ < ∞∫ ∫ ∫ .

However, if 2 2 4( ) then ( ) and ( )x t dt X f df X f df∞ ∞ ∞

−∞ −∞ −∞

< ∞ < ∞ < ∞∫ ∫ ∫ . This result also

applies to h(t).

( ) ( ) ( )Y f H f X f= 2 * *

2 2

22 4 4

2

( ) ( ) ( ) ( ) ( )

( ) ( )

( ) ( ) ( )

( )

Y f df X f H f X f H f df

X f H f df

Y f df X f df H f df

Y f df

∞ ∞

−∞ −∞

∞

−∞

∞ ∞ ∞

−∞ −∞ −∞

∞

−∞

= ⋅

=

≤

< ∞

∴ < ∞

∫ ∫

∫

∫ ∫ ∫

∫

By Rayleigh’s theorem: 2 2( ) ( )Y f df y t dt∞ ∞

−∞ −∞

=∫ ∫

2( )y t dt∞

−∞

∴ < ∞∫


2.17. The transfer function of the summing block is: [ ]1( ) 1 exp( 2 )H f j fTπ= − − .

The transfer function of the integrator is: 21( )

2H f

j fπ=

These elements are cascaded :

( ) ( )

( )[ ]

( )[ ]

1 2 1 2

22

2

( ) ( ) ( ) ( ) ( )1 1 exp( 2 )

21 1 2exp( 2 ) exp( 4 )

2

H f H f H f H f H f

j fTf

j fT j fTf

ππ

π ππ

= ⋅

= − − −

= − − − + −


2.18.a) Using the Laplace transform representation of a single stage, the transfer function is:

0

0

00

1( )1

1 1

1( )1 2

H sRCs

s

H fj f

τ

π τ

=+

=+

=+

These units are cascaded, so the transfer function for N stages is:

( )0

1( ) ( )1 2

NNH f H f

j fπ τ⎛ ⎞

= = ⎜ ⎟+⎝ ⎠

b) For N→∞, and 2

20 24

TN

τπ

=

( )0

0

1ln ( ) ln1 2

ln 1 2

ln 1

let , then for very large , 1

H f Nj f

N j f

jfTNN

jfTz N zN

π τ

π τ

⎛ ⎞= ⎜ ⎟+⎝ ⎠= − +

⎛ ⎞= − +⎜ ⎟⎝ ⎠

= <

We can use the Taylor series expansion of ln(1 )z∴ +

( )

( )

1

1

1

1

1ln(1 ) 1

1 1

m m

m

mm

m

N z N zm

fTN jm N

∞+

=

∞+

=

⎡ ⎤− + = − −⎢ ⎥⎣ ⎦⎡ ⎤⎛ ⎞= − −⎢ ⎥⎜ ⎟

⎝ ⎠⎢ ⎥⎣ ⎦

∑

∑

(next page)


2.18 (b) Cont’d Taking the limit as N→∞:

( )2 2

1

1

2 2

1lim 12

1 2

mm

N m

fT fT f TN j N jm NN N

f T j N fT

∞+

→∞=

⎛ ⎞⎡ ⎤ ⎛ ⎞⎛ ⎞⎜ ⎟− − = − +⎢ ⎥ ⎜ ⎟⎜ ⎟⎜ ⎟⎝ ⎠ ⎝ ⎠⎢ ⎥⎣ ⎦⎝ ⎠

= − −

∑

2 2

2 2

1( ) exp( ) exp( )21( ) exp( )2

H f f T j N ft

H f f T

∴ = − −

∴ = −


2.19.a) ( ) ( )T

t T

y t x dτ τ−

= ∫

This is the convolution of a rectangular function with x(τ). The interval of the rectangular function is [(t-T),T], and the midpoint is T/2.

Tsinc( ), but the function is shifted by .2

( ) sinc( ) exp( )

trect T fTT

H f T fT j fTπ

⎛ ⎞⎜ ⎟⎝ ⎠

∴ = −

b)BW = 1 1RC T

=

( ) exp( 2 )1 2 2

1 exp( )1 2

1( ) exp ( ) ( )2 2

1 exp ( ) ( )2 2

T TH f j fj RC f

T j fTRC j f

RCT T Th t t u t

RC RCT Tt u t

T

ππ

ππ

= −+

⎛ ⎞⎜ ⎟

= −⎜ ⎟⎜ ⎟+⎝ ⎠

⎛ ⎞∴ = − − −⎜ ⎟⎝ ⎠

⎛ ⎞= − − −⎜ ⎟⎝ ⎠


2.20. a) For the sake of convenience, let h(t) be the filter time-shifted so that it is symmetric about the origin (t = 0).

1 12 2

01 1

12

1

( ) exp( 2 ) exp( 2 )

2 cos(2 )

N N

k kk k

N

kk

H f w j fk w j fk w

w fk

π π

π

− −−

= =−

−

=

= − + − +

=

∑ ∑

∑

Let G(f) be the filter returned to its correct position. Then 1( ) ( ) exp( 2 )

2NG f H f j fπ −⎛ ⎞= − ⎜ ⎟

⎝ ⎠, which is a time-shift of 1

2N −⎛ ⎞

⎜ ⎟⎝ ⎠

samples.

( )( )1

2

1( ) exp 1 2 cos(2 )

N

kk

G f j f N w fkπ π

−

=

∴ = − − ∑

b)By inspection, it is apparent that:

( ) exp( ( 1))G f j f Nπ= − − This meets the definition of linear phase.


2.21 Given an ideal bandpass filter of the type shown in Fig P2.7, we need to find the response of the filter for 0( ) cos(2 )x t A f tπ=

[ ]0 0

1 1( ) rect rect2 2 2 21( ) ( ) ( )2

c cf f f fH fB B B B

X f f f f fδ δ

− +⎛ ⎞ ⎛ ⎞= +⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠

= − + −

If 0cf f− is large compared to 2B, then the response is zero in the steady state. However:

0 00 0

( ) ( ) ( ) ( )2 ( ) 2 2 ( ) 2

A A A Ax t u t f f f fj f f j f f

δ δπ π

⎛ ⎞+ − + + +⎜ ⎟− +⎝ ⎠

Since 0cf f− is large, assume that the portion of the amplitude spectrum lying inside the

passband is approximately uniform with a magnitude of 04 ( )c

Af fπ −

.

The phase spectum of the input is plotted as: The approximate magnitude and phase spectra of the output:


Taking the envelope by retaining the positive frequency components, shifting them to the origin, and scaling by 2:

0

0

exp 22 if ( )

2 ( )0 otherwise

c

A j j ftB f BY f

f f

π π

π

⎧ ⎛ ⎞⎛ ⎞− −⎜ ⎟⎪ ⎜ ⎟⎝ ⎠⎪ ⎝ ⎠ − < <⎨ −⎪⎪⎩

[ ]

[ ]

00

00

( ) sinc 2 ( )( )

( ) sinc 2 ( ) sin(2 )( )

c

cc

ABy t B t tj f f

ABy t B t t f tf f

π

ππ

= −−

∴ −−


2.22 ( ) ( ) exp( 2 )H f X f j fTπ= −

[ ]

[ ]

( ) ( ) ( ) sinc( )exp( 2 )2 2

sinc( ( )) sinc( ( )) exp( )2

c c

c c

A TX f f f f f T fT j f

AT T f f T f f j fT

δ δ π

π

= − + + ∗ −

= − + + −

Let for largecNf NT

=

( ) ( )

( ) ( ) ( ) ( )2 2

2 2

( ) ( ) ( )

( ) exp( 2 )exp( ) sinc ( ) sinc ( )2

exp( 2 ) sinc ( ) sinc ( ) sinc ( ) sinc ( )4

exp( 2 ) sinc( ) sinc(4

c c

c c c c

Y f H f X fATX f j fT j fT T f f T f f

A Tj fT T f f T f f T f f T f f

A Tj fT fT N f

π π

π

π

=

= − − − + +⎡ ⎤⎣ ⎦

= − + + − − + − +⎡ ⎤ ⎡ ⎤⎣ ⎦ ⎣ ⎦

= − − + −[ ][ ]) sinc( ) sinc( )T N fT N fT N+ − + +

But sinc(x)=sinc(-x)

[ ]2 2

( ) exp( 2 ) sinc( ) sinc( )2

A TY f j fT fT N fT Nπ∴ = − + +


2.23 G(k)=G

1

01

0

1

0

1 2( )exp( )

2 exp( )

2 2 cos( ) sin( )

N

nkN

k

N

k

g G k j k nN NG j k nN NG j k n j j k nN N N

π

π

π π

−

=

−

=

−

=

= ⋅

= ⋅

= ⋅ + ⋅

∑

∑

∑

If n = 0, 1

0( ) 1

N

k

Gg n GN

−

=

= =∑

For 0n ≠ , we are averaging over one full wavelength of a sine or cosine, with regularly sampled points. These sums must always be zero.


2.24. a) By the duality and frequency-shifting properties, the impulse response of an ideal low-pass filter is a phase-shifted sinc pulse. The resulting filter is non-causal and therefore not realizable in practice. c)Refer to the appropriate graphs for a pictorial representation. i)Δt=T/100 BT Overshoot (%) Ripple Period 5 9,98 1/5 10 9.13 1/10 20 9.71 1/20 100 100 No visible ripple



2.24 (d) Δt Overshoot (%) Ripple Period T/100 100 No visible ripple. T/150 16.54 1/100 T/200 ~0 No visible ripple. Discussion Increasing B, which also increases the filter’s bandwidth, allows for more of the high-frequency components to be accounted for. These high-frequency components are responsible for producing the sharper edges. However, this accuracy also depends on the sampling rate being high enough to include the higher frequencies.


2.25 BT Overshoot (%) Ripple Period 5 8.73 1/5 10 8.8 1/10 20 9.8 1/20 100 100 - The overshoot figures better for the raised cosine pulse that for the square pulse. This is likely because a somewhat greater percentage of the pulse’s energy is concentrated at lower frequencies, and so a greater percentage is within the bandwidth of the filter.




2.26.b)


2.26 b) If B is left fixed, at B=1, and only T is varied, the results are as follows BT Max. Amplitude 5 1.194 2 1.23 1 1.34 0.5 0.612 0.45 0.286 As the centre frequency of the square wave increases, so does the bandwidth of the signal (and its own bandwidth shifts its centre as well). This means that the filter passes less of the signal’s energy, since more of it will lie outside of the pass band. This results in greater overshoot. However, as the frequency of the pulse train continues to increase, the centre frequency is no longer in the pass band, and the resulting output will also be attenuated.


c) BT Max. Amplitude 5 1.18 2 1.20 1 1.27 0.5 0.62 0.45 0.042 Extending the length of the filter’s impulse response has allowed it to better approximate the ideal filter in that there is less ripple. However, this does not extend the bandwidth of the filter, so the reduction in overshoot is minimal. The dramatic change in the last entry (BT=0.45) can be accounted for by the reduction in ripple.


2.27 a)At fs = 4000 and fs = 8000, there is a muffled quality to the signals. This improves with higher sampling rates. Lower sampling rates throw away more of the signal’s high frequencies, which results in a lower quality approximation. b)Speech suffers from less “muffling” than do other forms of music. This is because a greater percentage of the signal energy is concentrated at low frequencies. Musical instruments create notes that have significant energy in frequencies beyond the human vocal range. This is particularly true of instruments whose notes have sharp attack times.


2.28



Chapter 3

3.1 s(t) = Ac[1+kam(t)]cos(2πfc t) where m(t) = sin(2πfs t) and fs=5 kHz and fc = 1 MHz.

( ) [cos(2 ) (sin(2 ( ) ) sin(2 ( ) )]2a

c c c c sks t A f t f fs t f f tπ π π∴ = + + + −

s(t) is the signal before transmission.

The filter bandwidth is: 610 5714 Hz

175cfBW

Q= = =

m(t) lies close to the 3dB bandwidth of the filter, m(t) is therefore attenuated by a factor of a half.

' '

'

( ) 0.5 ( ) or 0.5

0.25a a

a

m t m t k k

k

∴ = =

∴ =

The modulation depth is 0.25


3.2 (a)

0[exp( ) 1]T

vi IV

= − −

Using the Taylor series expansion of exp(x) up to the third order terms, we get:

2 3

01 1[ ]2 6T T T

v v vi IV V V

⎛ ⎞ ⎛ ⎞= − + −⎜ ⎟ ⎜ ⎟

⎝ ⎠ ⎝ ⎠

(b) ( ) 0.01[cos(2 ) cos(2 )]m cv t f t f tπ π= +

Let 2 , 22 2

c m c mf f f ft tθ π φ π+ −= =

then ( ) 0.02[cos cos ]v t θ φ=

2 2

2

2

( ) 0.02 [1 cos(2 )][1 cos(2 )]10.02 [1 cos(2 ) cos(2 ) (cos(2 2 ) cos(2 2 ))]2

10.02 [1 cos(2 ( ) ) cos(2 ( ) ) (cos(4 ) cos(4 ))]2c m c m c m

v t

f f t f f t f t f t

θ φ

θ φ θ φ θ φ

π π π π

∴ = + +

= + + + + + −

= + + + − + +

3 3

3

3cos cos3 3cos cos3( ) 0.024 4

0.02 9 3[ (cos( ) cos( )) (cos( 3 ) cos( 3 )16 2 2

3 1(cos(3 ) cos(3 )) (cos(3 3 ) cos(3 3 ))]2 2

v t θ θ φ φ

θ φ θ φ θ φ θ φ

θ φ θ φ θ φ θ φ

+ +⎡ ⎤ ⎡ ⎤= ⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦

= + + − + + + −

+ + + − + + + −

23 0.02 9 3( ) [ (cos(2 ) cos(2 )) (cos(2 (2 ) ) cos(2 (2 ) )

16 2 23 1(cos(2 (2 ) ) cos(2 (2 ) )) (cos(6 ) cos(6 ))]2 2

c m c m m t

c m m t c m

v t f t f t f f t f f t

f f t f f t f t f t

π π π π

π π π π

∴ = + + − + −

+ + + + + +


The output will have spectral components at: fm fc fc+ fm fc- fm 2fc 2fm 2fc- fm 2fc+ fm fc- 2fm fc+2 fm 3fc 3fm (c) The bandpass filter must be symmetric and centred around fc . It must pass components at fc+ fm, but reject those at fc+2 fm and higher.


(d) Term # Carrier Message Taylor Coef. 1 0.01 -38.46 2 0.0001 739.6 3 2.25 x 10-6 -9.48 x 103

After filtering and assuming a filter gain of 1, we get: ( ) 0.41cos(2 ) 0.074[cos(2 ( ) ) cos(2 ( ) )]0.41cos(2 ) .148[cos(2 )cos(2 )][0.41 0.148cos(2 )]cos(2 )[1 0.36cos(2 )]cos(2 )

The modulation percentage is ~36%

c c m c m

c c m

m c

m c

i t f t f f t f f tf t f t f t

f t f tf t f t

π π ππ π π

π ππ π

= + − + += +

= += +

∴










3.10. The circuit can be rearranged as follows: (a)

(b)

Let the voltage Vb-Vd be the voltage across the output resistor, with Vb and Vd being the voltages at each node. Using the voltage divider rule for condition (a):

, , = f b fbb d b d

f b f b f b

R R RRV V V V V V VR R R R R R

−= = −

+ + +

and for (b):

, , =f b fbb d b d

f b f b f b

R R RRV V V V V V VR R R R R R

− += − = − −

+ + +

∴The two voltages are of the same magnitude, but are of the opposite sign.






3.16 (a)

1 1( ) cos(2 ( ) ) (1 ) cos(2 ( ) )2 2

( ) [ (cos(2 )cos(2 ) sin(2 )sin(2 ))2

(1 )(cos(2 )cos(2 ) sin(2 )sin(2 ))]

( ) [cos(2 )cos(2 ) (1 2 )si2

m c m c m c m c

m cc m c m

c m c m

m cc m

s t a A A f f t a A A f f t

A As t a f t f t f t f t

a f t f t f t f t

A As t f t f t a

π π

π π π π

π π π π

π π

= ⋅ + + − +

= −

+ − +

= + −

1

2

n(2 )sin(2 ))]

( ) cos(2 )2

( ) (1 2 )sin(2 )2

c m

mm

mm

f t f t

Am t f t

Am t a f t

π π

π

π

∴ =

= −

b)Let:

1 1( ) ( )cos(2 ) ( )sin(2 )2 2c c c s cs t A m t f t A m t f tπ π= +

By adding the carrier frequency:

1 1( ) [1 ( )]cos(2 ) ( )sin(2 )2 2c a c a c s cs t A k m t f t k A m t f tπ π= + +

where ak is the percentage modulation. After passing the signal through an envelope detector, the output will be:

12 2 2

12 2

1 1( ) 1 ( ) ( )2 2

1 ( )1 2 1 ( ) 1 12 1 ( )2

c a a s

a s

c a

a

s t A k m t k m t

k m tA k m t

k m t

⎧ ⎫⎪ ⎪⎡ ⎤ ⎡ ⎤= + +⎨ ⎬⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦⎪ ⎪⎩ ⎭

⎧ ⎫⎡ ⎤⎪ ⎪⎢ ⎥⎪ ⎪⎡ ⎤= + ⋅ +⎨ ⎬⎢ ⎥⎢ ⎥⎣ ⎦ ⎪ ⎪⎢ ⎥+

⎣ ⎦⎪ ⎪⎩ ⎭

The second factor in ( )s t is the distortion term d(t). For the example in (a), this becomes:

12 21 (1 2 )sin(2 )

2( ) 1 11 cos(2 )2

m

m

a f td t

f t

π

π

⎧ ⎫⎡ ⎤−⎪ ⎪⎢ ⎥⎪ ⎪= +⎨ ⎬⎢ ⎥⎪ ⎪⎢ ⎥+

⎣ ⎦⎪ ⎪⎩ ⎭


c)Ideally, d(t) is equal to one. However, the distortion factor increases with decreasing a. Therefore, the worst case exists when a = 0.






3.20. m(t) contains {100,200,400} Hz

The transmitted SSB signal is: ˆ[ ( ) cos(2 ) ( )sin(2 )2

cc c

A m t f t m t f tπ π−

Demodulation is accomplished using a product modulator and multiplying by: ' 'cos(2 )c cA f tπ

(a)

' '1 ˆ( ) cos(2 )[ ( )cos(2 ) ( )cos(2 )]2o c c c c cv t A A f t m t f t m t f tπ π π= −

The only lowpass components will be those that are functions of only t and Δf. Higher frequency terms will be filtered out, and so can be ignored for the purposes of determining the output of the detector.

'1 ˆ( ) [ ( ) cos(2 ) ( )sin(2 )]4o c cv t A A m t f t m t f tπ π∴ = Δ − Δ by using basic trig identities.

When the upper side-band is transmitted, and Δf>0, the frequencies are shifted inwards by Δf.

( ) contains {99.98,199.98,399.98} HzoV f∴ (b) When the lower side-band is transmitted, and Δf>0, then the baseband frequencies are shifted outwards by Δf.

( ) contains {100.02,200.02,400.02} HzoV f∴




3.22.

1 2 1 2 1 1 2 2

1 21 2 1 2 1 2 1 2

( ) ( ) cos(2 )cos(2 )

[cos(2 ( ) ) cos(2 ( ) )]2

v t v t A A f t f tA A f f t f f t

π φ π φ

π φ φ π φ φ

= + +

= − + − + + + +

The low-pass filter will only pass the first term.

1 2 1 2 1 21( ( ) ( )) [cos( 2 ( 2 ) )]2

LFP v t v t A A W f tπ φ φ∴ = − + Δ + −

Let v0(t) be the final output, before band-pass filtering.

1 21 2 2 2 2

2 1 2 1 21 2 2 2 2

2 1 2 1 21 2 2 2

1 2( ) [cos( 2 ) cos(2 )]2 / 2 / 21 [cos( 2 ) cos(2 )]2 2 21 [cos( 2 ( 2 ) ) cos( 2 )]4 2 2

o

c c

W fv t A A t A f tW f W f

A A ft f tn n

A A f f f tn n

φ φπ π φ

φ φ φ φπ φ π φ

φ φ φ φπ φ π φ

⎛ ⎞ −+ Δ= − + ⋅ +⎜ ⎟Δ + Δ +⎝ ⎠

− −= − Δ + − ⋅ + +

+ +− −

= − + Δ + − + − + ++ +

After band-pass filtering, retain only the second term.

2 1 21 2 2

1( ) [cos( 2 )4 2o cv t A A f t

nφ φπ φ−

∴ = − + ++

1 2

2

2

12

02 2

rearranging and solving for :

1

n n

n

φ φ φ

φφφ

− + =+ +

= −+

(b) At the second multiplier, replace v2(t) with v1(t). This results in the following expression for the phase:

1 21

21

02 2

3

n n

n

φ φ φ

φφ

− + =+ +

=+

1

2

c

c

f f f Wf f f= −Δ −= + Δ


3.23. Assume that the mixer performs a multiplication of the two signals.

1

2

( ) {1,2,3,4,5,6,7,8,9} MHz( ) {100,200,300,400,500,600,700,800,900} kHz

y ty t

∈∈

This system essentially produces a DSB-SC signal centred around the frequency of y1(t). The lowest frequencies that can be produced are:

1 2 1 2

1 1 2

2 1 2

1( ) [cos(2 ( ) ) cos(2 ( ) )]2

1 MHz 0.9 MHz100 kHz 1.1 MHz

oy t f f t f f t

f f ff f f

π π= − + +

= − == + =

The highest frequencies that can be produced are:

1 1 2

2 1 2

9 MHz 8.1 MHz900 kHz 9.9 MHz

f f ff f f= − == + =

The resolution of the system is the bandwidth of the output signal. Assuming that no branch can be zeroed, the narrowest resolution occurs with a modulation frequency of 100 kHz. The widest bandwidth occurs when there is a modulation frequency of 900 kHz.


3.24 Given the presence of the filters, only the baseband signals need to be considered. All of the other product components can be discarded. (a) Given the sum of the modulated carrier waves, the individual message signals are extracted by multiplying the signal with the required carrier. For m1(t), this results in the conditions:

1 1

2 2

3 3

cos( ) cos( ) 0cos( ) cos( ) 0cos( ) cos( ) 0

i i

α βα βα β

α β π

+ =+ =+ =

∴ = ±

For the other signals:

2

1 1 1 1

2 1 2 1 2 1 2 1

3 1 3 1 3 1 3 1

3

1 2 1 2

3 2 3 2

( ) :cos( ) cos( ) 0 cos( ) cos( ) 0 ( ) ( )cos( ) cos( ) 0 ( ) ( )

Similarly:( ) :

( ) ( )( ) (

m t

m t

α β α β πα α β β α α β β πα α β β α α β β π

α α β β πα α β β

− + − = = ±− + − = − = − ±− + − = − = − ±

− = − ±− = −

4

1 3 1 3

2 3 2 3

)

( ) :( ) ( )( ) ( )

m t

π

α α β β πα α β β π

±

− = − ±− = − ±

(b) Given that the maximum bandwidth of mi(t) is W, then the separation between fa and fb must be | fa- fb|>2W in order to account for the modulated components corresponding to fa- fb.


3.25 b) The charging time constant is ( ) 1f sr R C sμ+ = The period of the carrier wave is 1/fc = 50 μs. The period of the modulating wave is 1/fm = 0.025 s. ∴The time constant is much shorter than the modulating wave and therefore should track the message signal very well. The discharge time constant is: 100lR C sμ= . This is twice the period of the carrier wave, and should provide some smoothing capability. From a maximum voltage of V0, the voltage Vc across the capacitor after time t = Ts is:

0 exp( )sc

l

TV VR C

= −

Using a Taylor series expansion and retaining only the linear terms, will result in the

linear approximation of 0 (1 )sC

l

TV VR C

= − . Using this approximation, the voltage will

decay by a factor of 0.94 from its initial value after a period of Ts seconds. From the code, it can be seen that the voltage decay is close to this figure. However, it is somewhat slower than what was calculated using the linear approximation. In a real circuit, it would also be expected that the decay would be slower, as the voltage does not simply turn off, but rather decreases over time.


3.25 c)

The output of a high-pass RC circuit can be described according to:

0

0

( ) ( )( ) ( ( ) ( ))

( )

c in

c

V t I t RQ t C V t V t

dQI tdt

== −

=

00

( ) ( )( ) indV t dV tV t RCdt dt

⎛ ⎞= −⎜ ⎟⎝ ⎠

Using first order differences to approximate the derivatives results in the following difference equation:

0 0( ) ( 1) ( ( ) ( 1))in ins s

RC RCV t V t V t V tRC T RC T

= − + − −+ +

The high-pass filter applied to the envelope detector eliminates the DC component.


Problem 3.25. MATLAB code function [y,t,Vc,Vo]=AM_wave(fc,fm,mi) %Problem 3.25 %Inputs: fc Carrier Frequency % fm Modulation Frequency % mi modulation index %Problem 3.25 (a) fs=160000; %sampling rate deltaT=1/fs; %sampling period t=linspace(0,.1,.1/deltaT); %Create the list of time periods y=(1+mi*cos(2*pi*fm*t)).*cos(2*pi*fc*t); %Create the AM wave %Problem 3.25 (b) %%%%Create the envelope detector%%%% Vc=zeros(1,length(y)); Vc(1)=0; %inital voltage for k=2:length(y) if (y(k)>(Vc(k-1))) Vc(k)=y(k); else


Vc(k)=Vc(k-1)-0.023*Vc(k-1); end end %Problem 3.25 (c) %%%Implement the high pass filter%%% %%This implements bias removal Vo=zeros(1,length(y)); Vo(1)=0; RC=.001; beta=RC/(RC+deltaT); for k=2:length(y) Vo(k)=beta*Vo(k-1)+beta*(Vc(k)-Vc(k-1)); end


Chapter 4 Problems









Problem 4.7.

( ) cos( ( ))( ) 2 ( )

c

c p

s t A tt f t k m t

θθ π

== +

Let β = 0.3 for m(t) = cos(2πfmt).

( ) cos(2 ( )) [cos(2 )cos( cos(2 )) sin(2 )sin( cos(2 ))]for small :cos( cos(2 )) 1sin( sin(2 )) cos(2 )

c c

c c m c m

m

m m

s t A f t m tA f t f t f t f t

f tf t f t

π βπ β π π β π

ββ πβ π β π

∴ = += −

( ) cos(2 ) sin(2 )cos(2 )

cos(2 ) [sin(2 ( ) ) sin(2 ( ) )2

c c c c

cc c c m c m

s t A f t A f t fmtAA f t f f t f f t

π β π π

π β π π

∴ = −

= − + + +







Problem 4.14.

22 1

( ) cos(2 sin(2 )) cos(2 ( ))

c c m

c c

v avs t A f t f t

A f t m tπ β ππ β

== += +

2

22

( )

cos (2 ( ))

cos(4 2 ( ))2

c

c

v a s t

a f t m ta f t m t

π β

π β

= ⋅

= ⋅ +

= ⋅ +

The square-law device produces a new FM signal centred at 2fc and with a frequency deviation of 2β. This doubles the frequency deviation.




4.17. Consider the slope circuit response: The response of |X1(f)| after the resonant peak is the same as for a single pole low-pass filter. From a table of Bode plots, the following gain response can be obtained:

1 2

1| ( ) |

1 B

X ff f

B

=−⎛ ⎞+ ⎜ ⎟

⎝ ⎠

Where fB is the frequency of the resonant peak, and B is the bandwidth. For the slope circuit, B is the filter’s bandwidth or cutoff frequency. For convenience, we can shift the filter to the origin (with 1( )X f as the shifted version).

1 2

13

2 2

1| ( ) |

1

| ( ) |

(1 )f kB

X ffB

d X f kdf B k=

=⎛ ⎞+ ⎜ ⎟⎝ ⎠

= −+


Because the filters are symmetric about the central frequency, the contribution of the second filter is identical. Adding the filter responses results in the slope at the central frequency being:

32 2

| ( ) | 2

(1 )f kB

d X f kdf B k=

= −+

In the original definition of the slope filter, the responses are multiplied by -1, so do this here. This results in a total slope of:

32 2

2

(1 )

k

B k+

As can be seen from the following plot, the linear approximation is very accurate between the two resonant peaks. For this plot B = 500, f1=-750, and f2=750.








Problem 4. 23


Problem 4.24 The amplitude spectrum corresponding to the Gaussian pulse 2 2( ) exp * [ / ]p t c c t rect t Tπ⎡ ⎤= −⎣ ⎦ is given by the magnitude of its Fourier transform.

( ) ( ) ( )

[ ]

2 2

2 2

exp /

exp sinc

P f c c t rect t T

c f c T fT

π

π

⎡ ⎤ ⎡ ⎤= − ⎣ ⎦⎣ ⎦

⎡ ⎤= −⎣ ⎦

F F

where we have used the convolution theorem Problem 4.25 The Carson rule bandwidth for GSM is ( )2TB f W= Δ + where the peak deviation is given by

1 2 / log(2) 0.752 4

fk cf B Bπ

πΔ = = =

With BT = 0.3 and T = 3.77 microseconds, the peak deviation is 59.7 kHz From Figure 4.22, the one-sided 3-dB bandwidth of the modulating signal is approximately 50 kHz. Combining these two results, the Carson rule bandwidth is

( )2 59.7 50

219.4 kHzTB = +

=

The 1-percent FM bandwidth is given by Figure 4.9 with 59.7 1.1950

fW

β Δ= = = . From the

vertical axis we find that 6TBf=

Δ , which implies BT = 6(59.7) = 358.2 kHz.


Problem 4.26. a)

Beta # of side frequencies 1 1 2 2 5 8 10 14 b)By experimentation, a modulation index of 2.408, will force the amplitude of the carrier to be about zero. This corresponds to the first root of J0(β), as predicted by the theory.


Problem 4.27. a)Using the original MATLAB script, the rms phase error is 6.15 % b)Using the plot provided, the rms phase error is 19.83% Problem 4.28 a)The output of the detected signal is multiplied by -1. This results from the fact that m(t)=cos(t) is integrated twice. Once to form the transmitted signal and once by the envelope detector. In addition, the signal also has a DC offset, which results from the action of the envelope detector. The change in amplitude is the result of the modulation process and filters used in detection.

b)If ( ) sin(2 ) 0.5cos 23m

mfs t f t tπ π⎛ ⎞= + ⎜ ⎟

⎝ ⎠, then some form of clipping is observed.


The above signal has been multiplied by a constant gain factor in order to highlight the differences with the original message signal. c)The earliest signs of distortion start to appear above about fm =4.0 kHz. As the message frequency may no longer lie wholly within the bandwidth of either the differentiator or the low-pass filter. This results in the potential loss of high-frequency message components.


4.29. By tracing the individual steps of the MATLAB algorithm, it can be seen that the resulting sequence is the same as for the 2nd order PLL.

( ) is the phase error ( ) in the theoretical model.ee t tφ The theoretical model of the VCO is:

20

( ) 2 ( )t

vt k v t dtφ π= ∫

and the discrete-time model is: VCOState VCOState 2 ( 1)v sk t Tπ= + − which approximates the integrator of the theoretical model. The loop filter is a PI-controller, and has the transfer function:

( ) 1 aH fjf

= +

This is simply a combination of a sum plus an integrator, which is also present in the MATLAB code:

Filterstate Filterstate ( ) Integrator( ) Filterstate ( ) Integrator +input

e tv t e t

= += +

b)For smaller kv, the lock-in time is longer, but the output amplitude is greater.


c)The phase error increases, and tracks the message signal.

d)For a single sinusoid, the track is lost if 0 0 where m f v c vf K K k k A A≥ = For this question, K0=100 kHz, but tracking degrades noticeably around 60-70 kHz. e)No useful signal can be extracted. By multiplying s(t) and r(t), we get:

sin( VCOState) sin(4 VCOState)2c v

f c fA A k f t kφ π φ⎡ ⎤− + + +⎣ ⎦

This is substantially different from the original error signal, and cannot be seen as an adequate approximation. Of particular interest is the fact that this equation is substantially more sensitive to changes in φ than the previous one owing to the presence of the gain factor kv



Chapter 5 Problems

5.1. (a) Given 2

22

( )1( ) exp( )22

x

xx

xf x μσπσ

−= −

and 2 2exp( ) exp( )t fπ π− − , then by applying the time-shifting and scaling properties:

2 2 2 2

2

1( ) 2 exp( ( 2 ) )exp( 2 )2

x x x

x

F f f j fπσ π πσ π π μπσ

= −

= 2 2 2exp( 2 2 ) and let 2x xf j f fπ σ μ π ν π− + =

= 2 21exp( )2x xjνμ ν σ−

(b)The value of μx does not affect the moment, as its influence is removed. Use the Taylor series approximation of φx(x), given μx = 0.

2 2

2

0

1( ) exp( )2

exp( )!

x x

n

xxn

φ ν ν σ

∞

=

= −

=∑

0

2 2

0

( )[ ]

1 ( )2 !

nn x

nv

k k kx

xk

dE Xd

k

φ νν

σ νφ ν

=

∞

=

=

⎛ ⎞∴ = −⎜ ⎟⎝ ⎠

∑

For any odd value of n, taking ( )nx

n

ddφ νν

leaves the lowest non-zero derivative as ν2k-n.

When this derivative is evaluated for v=0, then [ ]nE X =0. For even values of n, only the terms in the resulting derivative that correspond to ν2k-n = ν0 are non-zero. In other words, only the even terms in the sum that correspond to k = n/2 are retained.

2! [ ]( / 2)!

nx

nE Xn

σ∴ =


5.2. (a) All the inputs for x ≤0 are mapped to y = 0. However, the probability that x > 0 is unchanged. Therefore the probability density of x ≤0 must be concentrated at y=0.

(b) Recall that ) 1 where ( ) is an even function.x xf x dx f x∞

−∞

=∫ Because fy(y) is a

probability distribution, its integral must also equal 1.

0 0

( ) 0.5 and ( ) 0.5x yf x dx f y dy+

∞ ∞

∴ = =∫ ∫

Therefore, the integral over the delta function must be 0.5. This means that the factor k must also be 0.5.


5.3 (a)

(b) ( ) ( )yP y p y dyα

α∞

≥ = ∫

Use the cumulative Gaussian distribution,

2

2

2, 2

1 ( )( ) exp( )22

y yy dyμ σ

μσπσ−∞

−Φ = −∫

2 21, 1,

1 ( ) [ ( ) ( )]2

P yσ σ

α α α−

∴ ≥ = Φ − +Φ −

But, 2,

1( ) [1 ( )]2 2

yy erfμ σ

μσ−

Φ = +

1 1 1 ( ) [2 ]2 2 2

P y erf erfα αασ σ− + − −⎛ ⎞ ⎛ ⎞∴ ≥ = + +⎜ ⎟ ⎜ ⎟

⎝ ⎠ ⎝ ⎠

0 0 1 1

0 1

0 1

2 2

2 22

( ) ( | ) ( ) ( | ) ( )

Assume: ( ) ( ) 0.51 ( ) [ ( | ) ( | )2

1 ( 1) ( 1)( ) [exp( ) exp( )]2 22 2

y y y

y y y

y

p y p y x P x p y x P x

P x P x

p y p y x p y y

y yp yσ σπσ

= +

= =

∴ = +

+ −= − + −


Problem 5.4


Problem 5.5 If, for a complex random process Z(t) [ ]( ) *( ) ( )ZR Z t Z tτ τ= +E then (i) The mean square of a complex process is given by

[ ]2

(0) *( ) ( )

( )

ZR Z t Z t

Z t

=

⎡ ⎤= ⎣ ⎦

E

E

(ii) We show ( )ZR τ has conjugate symmetry by the following

[ ][ ][ ]*

( ) *( ) ( )

*( ) ( )

( ) ( ) *

( )

Z

Z

R Z t Z t

Z s Z s

Z s Z s

R

τ τ

τ

τ

τ

− = −

= +

= +

=

E

E

E

where we have used the change of variable s = t - τ. (iii) Taking an approach similar to that of Eq. (5.67)

( )

( )( )[ ]

[ ] [ ]

[ ]{ }{ }

2

2 2

2

0 ( ) ( )

( ) ( ) *( ) *( )

( ) *( ) ( ) *( ) *( ) ( ) ( ) *( )

( ) ( ) *( ) *( ) ( ) ( )

2 ( ) 2 Re *( ) ( )

2 (0) 2Re ( )Z Z

Z t Z t

Z t Z t Z t Z t

Z t Z t Z t Z t Z t Z t Z t Z t

Z t Z t Z t Z t Z t Z t

Z t Z t Z t

R R

τ

τ τ

τ τ τ τ

τ τ τ

τ

τ

⎡ ⎤≤ ± +⎢ ⎥⎣ ⎦⎡ ⎤= ± + ± +⎣ ⎦

= ± + ± + + + +

⎡ ⎤ ⎡ ⎤= ± + ± + + +⎣ ⎦ ⎣ ⎦⎡ ⎤= ± +⎣ ⎦

= ±

E

E

E

E E E E

E E

Thus { }Re ( ) (0)Z ZR Rτ ≤ .


Problem 5.6 (a)

*1 2

1 1 1 2 1 2 1 2 1 2 2 2

[ ( ) ( )][( cos(2 ) cos(2 )) ( cos(2 ) cos(2 ))]

E Z t Z tE A f t jA f t A f t jA f tπ θ π θ π θ π θ= + + + ⋅ + + +

Let ω1=2πf1 ω2=2πf2 After distributing the terms, consider the first term:

21 1 1 1 2 1

2

1 1 2 1 1 2 1

[cos( )cos( )]

[cos( ( )) cos( ( ) 2 )]2

A E t t

A E t t t t

ω θ ω θ

ω ω θ

+ +

= − + + +

The expectation over θ1 goes to zero, because θ1 is distributed uniformly over [-π,π]. This result also applies to the term 2

2 1 2 2 2 2[cos( ) cos( )]A t tω θ ω θ+ + . Both cross-terms go to zero.

2

1 2 1 1 2 2 1 2 ( , ) [cos( ( )) cos( ( ))]2AR t t t t t tω ω∴ = − + −

(b) If f1 = f2, only the cross terms may be different:

21 1 2 1 2 1 1 1 2 1 2 1[ (cos( ) cos( ) cos( ) cos( )]E jA t t t tω θ ω θ ω θ ω θ+ + + + +

But, unless θ1=θ2, the cross-terms will also go to zero. 2

1 2 1 1 2 ( , ) cos( ( ))R t t A t tω∴ = − (c) If θ1=θ2, then the cross-terms become:

2 21 1 2 2 1 1 2 2 1 2 1 1 2 1 1 2 2 1[cos(( )) cos(( ) 2 ) [cos(( )) cos(( ) 2 )]jA E t t t t jA E t t t tω ω ω ω θ ω ω ω ω θ− − + + + + − + + +

After computing the expectations, the cross-terms simplify to:

2

2 1 1 2 1 1 2 2[cos( ) cos( )]2

jA t t t tω ω ω ω− − −

2

1 2 1 1 2 2 1 2 2 1 1 2 1 1 2 2 ( , ) [cos( ( )) cos( ( )) cos( ) cos( )]2ZAR t t t t t t j t t j t tω ω ω ω ω ω∴ = − + − + − − −


Problem 5.7


Problem 5.8


Problem 5.9



Problem 5.10


Problem 5.11


Problem 5.12


Problem 5.13


Problem 5.14


Problem 5.15


Problem 5.16


Problem 5.17


Problem 5.18


Problem 5.19

Problem 5.20



Problem 5.21




Problem 5.22



Problem 5.23


Problem 5.24


Problem 5.25

Problem 5.26


Problem 5.27


Problem 5.28

c)For a given filter, ( )H f , let ln ( )H fα =

and the Paley-Wiener criterion for causality is: 2

( )1 (2 )

fdf

fαπ

∞

−∞

< ∞+∫

For the filter of part (b)

[ ]01( ) ln(2) ln( ( ) ln( )2 xf S f Nα = + −

The first and the last terms have no impact on the absolute integrability of the previous expression, and so do not matter as far as evaluating the above criterion. This leaves the only condition:

2

ln ( )1 (2 )

xS fdf

fπ

∞

−∞

< ∞+∫


Problem 5.29


Problem 5.30


Problem 5.31



Problem 5.32


Problem 5.33

(a) The receiver position is given by x(t) = x0+vt Thus the signal observed by the receiver is

0

0

( , ) ( ) cos 2

( ) cos 2

( ) cos 2

c

c

cc c

xr t x A x f tc

x vtA x t f tc

f v xA x f t fc c

π

π

π

⎡ ⎤⎛ ⎞= −⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦⎡ ⎤+⎛ ⎞= −⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦⎡ ⎤⎛ ⎞= − −⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦

The Doppler shift of the frequency observed at the receiver is cD

f vfc

= .

(b) The expectation is given by

( ) ( )

( )

( )0

1exp 2 exp 2 cos2

1 exp 2 sin2

2

n D n n

D n n

D

j f j f d

j f d

J f

π

π

π

π

π τ π τ ψ ψπ

π τ ψ ψπ

π τ

−

−

⎡ ⎤ =⎣ ⎦

=

=

∫

∫

E

where the second line comes from the symmetry of cos and sin under a -π/2 translation.

Eq. (5.174) follows directly from this upon noting that, since the expectation result is real-valued, the right-hand side of Eq.(5.173) is equal to its conjugate.


Problem 5.34 The histogram has been plotted for 100 bins. Larger numbers of bins result in larger errors, as the effects of averaging are reduced. Distance Relative Error 0σ 0.94% 1σ 2.6 % 2σ 4.8 % 3σ 47.4% 4σ 60.7% The error increases further out from the centre. It is also important to note that the random numbers generated by this MATLAB procedure can never be greater than 5. This is very different from the Gaussian distribution, for which there is a non-zero probability for any real number.


5.34 Code Listing %Problem 5.34 %Set the number of samples to be 20,000 N=20000 M=100; Z=zeros(1,20000); for i=1:N for j=1:5 Z(i)=Z(i)+2*(rand(1)-0.5); end end sigma=sqrt(var(Z-mean(Z))); %Calculate a histogram of Z [X,C]=hist(Z,M); l=linspace(C(1),C(M),M); %Create a gaussian function with the same variance as Z G=1/(sqrt(2*pi*sigma^2))*exp(-(l.^2)/(2*sigma^2)); delta2=abs(l(1)-l(2)); X=X/(20000*delta2);


5.35 (a) For the generated sequence:

2

ˆ 0.0343 0.0493

ˆ 5.597y

y

jμ

σ

= − +

=

The theoretical values are: μy = 0 (by inspection). The theoretical value of 2

yσ =5.56. See 5.35 (c) for the calculation. 5.35 (b) From the plots, it can be seen that both the real and imaginary components are approximately Gaussian. In addition, from statistics, the sum of tow zero-mean Gaussian signals is also Gaussian distributed. As a result, the filter output must also be Gaussian.


5.35 (c)

Rh(z) = H(z)H(z-1) = But, Ry(z) = Rh(z)Rw(z) Taking the inverse z-transform:

2

2( ) 1

nwyr n a n

aσ

= −∞ < < ∞−

From the plots, the measured and observed autocorrelations are almost identical.

1

1

( ) ( 1) ( )( ) ( )

1( ) ( ) ( )1

n

y n ay n w nY z aY z z

H z h n a u naz

−

−

= − +

=

∴ = =−

1

1

2 1 2

1(1 )(1 )

1 11 1 1 1

az aza za az a az

−

−

−

− −

= +− − − −



Chapter 6 Solutions

Problem 6.1 The transfer function of the filter can be readily found to be

1( )1 2

H fj fRCπ

=+

Likewise, the power spectral density of the filtered noise is

02

/ 2( )1 (2 )N

NS ffRCπ

=+

Let a=1/RC,

02 2

/ 2( )(2 )N

aNS fa fπ

=+

Therefore, noting that

2 2

2exp( )(2 )

aaa f

τπ

−+

then the autocorrelation function of the output noise is

0( ) exp( )4NNRRC RC

ττ = −

For a zero-mean noise signal, the output power is simply RN(0), which is 0

4NRC

The output of the filtered sinusoid is:

1( ) [ ( ) ( )]2 1 2c cAS f f f f f

j fRCδ δ

π= − + + ⋅

+

And the resulting output power is

( )

2

21

2 1 2A

fRCπ+


Therefore the SNR is: 2

10 20

4 110 log dB1 (2 )

RCAN fRCπ

⋅+

Problem 6.2 The transfer function of the circuit can be found to be

( ) 122

RH fR j fL

j fCπ

π

=+ +

where 12cf LCπ

= and 1 LQR C

=

1( )

1 [( / ) ( / )]c c

H fjQ f f f f

∴ =+ −

For 1Q , the transfer function may be approximated as follows:

1 01 2 ( ) /

( )1 0

1 2 ( ) /

c c

c c

fj Q f f f

H ff

j Q f f f

⎧ >⎪ + −⎪= ⎨⎪ <⎪ + +⎩

For a noise source with a PSD of N0/2, the PSD of the filtered output will be

02 2 2

02 2 2

/ 2 01 4 ( ) /

( )/ 2 0

1 4 ( ) /

c cN

c c

N fQ f f f

S fN f

Q f f f

⎧ >⎪ + −⎪= ⎨⎪ >⎪ + +⎩

which is a symmetric function about the f = 0 axis. However,

( ) ( ) ( ) ( )NI NQ N c N cS t S t S f f S f f= = − + + Around f = fc, this allows us to approximate the PSDs of the in-phase and quadrature components as follows


02( ) ( )

1 (2 / )NI NQc

NS f S fQf f

=+

The variance of the in-phase and quadrature components of n(t) however, are the same as the variance of n(t) itself. Therefore, by taking the inverse Fourier transform of the above PSD and setting τ=0, we obtain

1( ) exp( )21(0)2

c cN

cN

f fRQ QfR

Q

π πτ τ

π

= −

=

which is the approximate variance (power) of the narrow band noise. Therefore, the SNR is,

2

1010 log dBc

A Qfπ

⋅


Problem 6.3



Problem 6.4


Problem 6.5



Problem 6.6 Problem 6.7


Problem 6.8





Problem 6.11


Problem 6.12


Problem 6.13


Problem 6.24



Problem 6.15


Problem 6.16 The following Matlab script simulates the generation and detection of an AM-modulated signal in noise.

%----------------------------------------- % Matlab code for Problem 6.16 %---------------------------------------- function Prob6_16() Fs = 143; % sample rate (kHz) t = [0: 1/Fs : 100]; % observation period (ms) Fc = 20; % carrier frequency (kHz) Fm = 0.1; % modulation frequency (kHz) Ka = 0.5; % modulation index SNRc = 25; % Channel SNR (dB) Ac = 1; tau = 0.25/4; %------------------------------------- % Modulated signal %------------------------------------- m = cos(2*pi*fm*t); C = Ac*cos(2*pi*fc*t); s = (1 + ka*m).* c; subplot(4,1,1), plot(t,s), grid on P = std(m)^2; %---------------------------------------------------- % Add narrowband noise % Create bandpass noise by low-pass % filtering AWGN noise and converting to % bandpass %------------------------------------------------------------ P_AM = Ac^2*(1+ka^2*P)/2; N = P_AM/10.^(SNRc/10); sigma = sqrt(N); %--- Create bandpass noise by low-pass filtering complex noise --- noise = randn(size(s)) + j*randn(size(s)); LPFnoise = LPF(Fs, noise, tau); BPnoise = real(LPFnoise .* exp(j*2*pi*fc/Fs*[1:length(s)])); scale = 2*sigma / std(BPnoise); s_n = s + scale * BPnoise; subplot(4,1,2), plot(t,s_n), grid on %--- Envelope detection of both noisy and noise-free signals --- ED = EnvDetector(t,s); ED_n = EnvDetector(t,s_n); %--- Remove transient and dc --- ED = ED(400:end); ED_n = ED_n(400:end); t = t(400:end); ED = ED - mean(ED); ED_n = ED_n - mean(ED_n); %--- Low pass filter ---- BBsig = LPF(Fs,ED,tau); BBsig_n = LPF(Fs,ED_n,tau);


%--- plot results ------ subplot(4,1,3), plot(t,BBsig); subplot(4,1,4), plot(t,BBsig_n) %--------------------------------------- % Envelope Detector from Problem 3.25 %--------------------------------------- function Vc = EnvDetector(t,s); Vc(1) = 0; % initial capacitor voltage for i = 2:length(s) if s(i) > Vc(i-1) % diode on Vc(i) = s(i); else % diode off Vc(i) = Vc(i-1) - 0.023*Vc(i-1); end end % plot(t, Vc), grid on return; %--------------------------------------- % Low pass filter %--------------------------------------- function y = LPF(Fs, x, tau); % tau = 1; % time constant of RC filter (ms) t1 = [0: 1/Fs : 5*tau]; h = exp(-t1/tau) * 1/Fs; y = filter(h, 1, x); return;

The Matlab script produces the following plot:


0 10 20 30 40 50 60 70 80 90 100-2

0

2

0 10 20 30 40 50 60 70 80 90 100-2

0

2

0 10 20 30 40 50 60 70 80 90 100-0.05

0

0.05

0 10 20 30 40 50 60 70 80 90 100-0.05

0

0.05

Time (ms)

Figure 6.16 Plot from Matlab script (a) AM modulated carrier (b) AM modulated carrier plus noise (c) AM demodulated signal in absence of noise (d) AM demodulated signal in noise


Problem 6.17 The following Matlab script simulates the generation and detection of an FM modulated signal in noise.

%----------------------------------------- % Problem 6.17 %----------------------------------------- function b = Prob6_17; %--- Parameters ------------------- fc = 100; % Carrier frequency (kHz) Fs = 1024; % Sampling rate (kHz) fm = 0.5; % Modulating frequency (kHz) Ts = 1/Fs; % Sample period (ms) t = [0:Ts:10]; % Observation period (ms) C_N = 20 % channel SNR (dB) Ac = 1; Bt = 20 % (kHz) W = 5; % (kHz) SNRc = C_N+10*log10(Bt/W); %--- Message signal --------------- m = cos(2*pi*fm*t); % modulating signal kf = 2.4; % modulator sensitivity index (~Bt/2) (kHz/V) %--- FM modulate ---------------- FMsig = FMmod(fc,t,kf,m,Ts); %--- Add narrowband noise --------- %--- Create bandpass noise by low-pass filtering complex noise --- P = 0.5; N = P/10.^(SNRc/10); sigma = sqrt(N); noise = randn(size(FMsig)) + j*randn(size(FMsig)); LPFnoise = LPF(Fs, noise, 0.05); % 0.01 => Bt ` 50 kHz eq. Noise BW BPnoise = real(LPFnoise .* exp(j*2*pi*fc/Fs*[1:length(FMsig)])); scale = sigma / std(BPnoise); FMsign = FMsig + scale * BPnoise; subplot(4,1,1), plot(t,FMsig), grid on subplot(4,1,2), plot(t,FMsign), grid on %--- FM receiver ---- Rx_c = FMdiscriminator(fc,FMsig,Ts); Rx_n = FMdiscriminator(fc,FMsign,Ts); t = t(round(1/Ts):end); % remove transient subplot(4,1,3), plot(t,Rx_c); grid on subplot(4,1,4), plot(t,Rx_n); grid on %--- Plot result --------- % FFTsize = 4096; % S = spectrum(FMsig,FFTsize); % % Freq = [0:Fs/FFTsize:Fs/2];


% subplot(2,1,1), plot(t,s), xlabel('Time (ms)'), ylabel('Amplitude'); % axis([0 0.5 -1.5 1.5]), grid on % subplot(2,1,2), stem(Freq,sqrt(S/682)), xlabel('Frequency (kHz)'), ylabel('Amplitude Spectrum'); % axis([95 105 0 1]), grid on %------------------------------------------------ % FM modulator %------------------------------------------------ function s = FMmod(fc,t,kf,m,Ts); theta = 2*pi*fc*t+ 2*pi*kf * cumsum(m)*Ts; % integrate signal s = cos(theta); %------------------------------------------------ % FM discriminator %------------------------------------------------ function D3 = FMdiscriminator(fc,S, Ts) t = [0:Ts:10*Ts]; % for filter %--- FIR differentiator (Fs = 1024 kHz, BT/2 = 10 kKhz) --- FIRdiff = [ 1.60385 0.0 0.0 0.0 -0.0 0.0 0.0 -0.0 -0.0 -0.0 -1.60385]; BP_diff = real(FIRdiff .* exp(j*2*pi*fc*t)); %--- Lowpass filter - Fs = 1024 kHz, f3dB = 5 kHz ----- LPF_B = 1E-4 *[ 0.0706 0.2117 0.2117 0.0706]; LPF_A = [1.0000 -2.9223 2.8476 -0.9252]; D1 = filter(BP_diff, 1, S); % Bandpass discriminator D2 = EnvDetect(D1); % Envelope detection D2 = D2 - mean(D2); % remove dc D3 = filter(LPF_B,LPF_A, D2); % Low-pass filtering D3 = D3(round(1/Ts):end); % remove transient (approx 1s) %--------------------------------------- % Envelope Detector %--------------------------------------- function Vc = EnvDetect(s); Vc(1) = 0; % initial capacitor voltage for i = 2:length(s) if s(i) > Vc(i-1) % diode on Vc(i) = s(i); else % diode off Vc(i) = Vc(i-1) - 0.005*Vc(i-1); end end return; %--------------------------------------- % Low pass filter %--------------------------------------- function y = LPF(Fs, x, tau); % tau = 1; % time constant of RC filter (ms) t1 = [0: 1/Fs : 5*tau]; h = exp(-t1/tau) * 1/Fs; y = filter(h, 1, x); return;


The Matlab script produces the following plot:

0 1 2 3 4 5 6 7 8 9 10-1

0

1

0 1 2 3 4 5 6 7 8 9 10-2

0

2

0 1 2 3 4 5 6 7 8 9 10-0.2

0

0.2

0 1 2 3 4 5 6 7 8 9 10-0.2

0

0.2

Time (ms)

Figure 6.17 Plot from Matlab script (a) FM modulated carrier (b) FM modulated carrier plus noise (c) FM demodulated signal in absence of noise (d) FM demodulated signal in noise


Chapter 8

Problem 8.1









Problem 8.4


Problem 8.5





Problem 8.6


Problem 8.7



Problem 8.8

Problem 8.9


Problem 8.10


Problem 8.11

Problem 8.12

. Problem 8.13

Problem 8.14

Problem 8.15


Problem 8.16

Problem 8.17

(a) The worst case ISI occurs if all preceding pulses have the same polarity

In this case, the received signal is

1

0

( ) ( )

( ) ( )

ii

i

r t a p t iT

a p t p t iT

∞

=−∞

−

=−∞

= −

= + −

∑

∑

where we have used the fact that the pulse is one sided. Substituting the pulse shape in the summation, we obtain

[ ]

1

0

01

( ) ( ) exp

( ) exp exp

i

i

t iTr t a p tT

ta p t iT

−

=−∞

∞

=

−⎡ ⎤= + −⎢ ⎥⎣ ⎦⎡ ⎤= + − −⎢ ⎥⎣ ⎦

∑

∑

Recognizing that this is a geometric summation, we obtain 1

0 1

0

0

( ) ( ) exp1

1( ) exp1

( ) 0.582 ( )

t er t a p tT eta p tT e

a p t p t

−

−⎡ ⎤= + −⎢ ⎥ −⎣ ⎦⎡ ⎤= + −⎢ ⎥ −⎣ ⎦

= +

In this example the ISI is nearly 60% of the original pulse. (b) If the time constant τ is not equal to T, then the calculation of the part (a) can be repeated for the generic case. In which case, we obtain the following expression

( )0

1( ) ( ) expexp / 1

tr t a p tTτ τ

⎡ ⎤= + −⎢ ⎥ −⎣ ⎦

If the maximum reduction of the eye opening is 20%, then by solving

10.20exp( / ) 1T τ

=−

we find that τ = T/ln(6) = 0.558T. Recall from Chapter Example 2.2, that the spectrum of the one-sized exponential pulse is


2

2( )1 (2 )

S ff

τπ τ

=+

and thus the 3-dB bandwidth is B3dB = 1/2πτ. By decreasing τ to 55.8% of T, we have increased the 3-dB bandwidth of the signal by the inverse of this amount, in order to keep the ISI to a manageable amount. Problem 8.18 Since the analog frequency response of the system including the matched filter P(f) is given by ( ) exp ( )H f f T P f⎡ ⎤= −⎣ ⎦ (1) the aliased version of this frequency response is given by

( ) expan

n nH f f T P fT T

∞

=−∞

⎡ ⎤ ⎛ ⎞= − + +⎜ ⎟⎢ ⎥ ⎝ ⎠⎣ ⎦∑ for 2f

T< . (2)

For zero ISI we must have Ha(f) = 1 for 12

fT

< . There are many solutions to this

problem. To reduce the number of options, we simply choose a function H(f) that satisfies

Ha(f) = 1 for 12

fT

< . One such function is

11( )

10

f T fTH f

fT

⎧ − <⎪⎪= ⎨⎪ >⎪⎩

(3)

Then we solve Eq.(1) for P(f) and obtain

( ) 11 exp

( )10

f T f T fTP f

fT

⎧ ⎡ ⎤− <⎣ ⎦⎪⎪= ⎨⎪ >⎪⎩

(4)


Problem 8.19 (a) The time domain response of the trapezoidal pulse is the inverse Fourier transform of the frequency domain response

[ ]

[ ] [ ] [ ]0.5 0.5 1.5

1.5 0.5 0.5

( ) ( ) exp 2

1.5 1.5exp 2 1exp 2 exp 2W W W

W W W

p t P f j ft df

f W W fj ft df j ft df j ft dfW W

π

π π π

∞

−∞

−

− −

=

+ −= + +

∫

∫ ∫ ∫

Noting the symmetry, we write this as

[ ] [ ] [ ]{ }

( ) ( ) ( ) ( )

( ) ( )

0.5 1.5

0.5 0.5

0.5 1.5 1.5 1.5

0.50.5 0.5 0.5

1.5( ) exp 2 exp 2 exp 2

2sin 2 2sin 2 2sin 2 2sin 21.5

2 2 2 2

sin sin 32 1.5

W W

W W

W W W W

WW W W

W fp t j ft df j ft j ft dfW

ft ft ft ftf dft t W t Wt

Wt Wtt

π π π

π π π ππ π π π

π ππ

−

−

−= + + −

⎡ ⎤⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎢ ⎥= + − −⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥⎣ ⎦ ⎣ ⎦ ⎣ ⎦⎣ ⎦

−= +

∫ ∫

∫

( ) ( ) ( )

( ) ( ) ( ) ( )

1.5

0.5

2 2

sin 1.5sin 3 0.5sin cos(2 )2 (2 )

sin sin cos 3 cos2

4

W

W

Wt Wt Wt ftt t Wt t

Wt Wt Wt Wtt t Wt

π π π ππ π π π

π π π ππ π π

⎡ ⎤ ⎡ ⎤− ⎡ ⎤− −⎢ ⎥ ⎢ ⎥ ⎢ ⎥

⎣ ⎦⎣ ⎦ ⎣ ⎦−

= − −

using the identity 2sinAsinB = cos(A-B)-cos(A+B), we rewrite the last line as

( ) ( ) ( )2 2

sin sin 2 sin( )

2Wt Wt Wt

p tt Wtπ π ππ π

= +

At the zero crossings of t = 1 1 3, 2 , 3 ,2 2b b bT T TW W W

± = ± ± = ± ± = ± , we have that p(t)

is zero. (b) Another pulse spectrum satisfying Nyquist criteria is the following

0 W/2-W/2 fW-W

P(f)


Problem 8.20






Problem 8.21



Problem 8.22



Problem 8.23


Problem 8.24 The provided Matlab script uses an approximation to the telephone model shown in Figure 8.9, but not exactly the same model used in Example 8.2. The response when using a 1.6 kbps NRZ signal is shown below. The response shows the same qualitative response as in Example 8.2 with level droop because the channel does not pass dc.

0 20 40 60 80 100 120 140 160 180 200

-1.5

-1

-0.5

0

0.5

1

1.5

Time samples

Am

plitu

de

Figure 8.24a-1 Response with NRZ signal.

The modified script for simulating the Manchester code is the following: %-------------------------------------------- % Problem 8.24 Telephone channel with Manchester code %-------------------------------------------- Fs = 32; % sample rate (kHz) Rs = 1.6; % symbol rate (kHz) Ns = Fs/Rs; % samples per symbol Nb = 30; % number of bits to simulate %--- Discrete B(z)/A(z) model of telephone channel --- A = [1.00, -2.838, 3.143, -1.709, 0.458, -0.049]; B = 0.1*[1.0, -1.0]; %---------------------------------------------------------------- % Simulate performance %================================================================ % pulse = [ones(1,Ns)]; % bipolar NRZ pulse pulse = [ones(1,Ns/2) -ones(1,Ns/2)]; % Manchester line code data = sign(randn(1,Nb)); Sig = pulse' * data; Sig = Sig(:); %--- Pass signal through channel ---- RxSig = filter(B,A,Sig); %--- Plot results ------------------------ plot(real(RxSig))


hold on, plot(Sig,'r'), hold off xlabel('Time samples'),ylabel('Amplitude') axis([0 200 -1.75 1.75]) % don't plot all samples

The signal output before and after the telephone line with a 1.6 kbps Manchester code is the following:

0 20 40 60 80 100 120 140 160 180 200

-1.5

-1

-0.5

0

0.5

1

1.5

Time samples

Am

plitu

de

Figure 8.24a-2 Response with Manchester code at 1.6 kbps.

The output shows much less level droop because the individual pulses do not have a dc component. If we increase the data rate to 3.2 kbps we obtain the following response:

0 20 40 60 80 100 120 140 160 180 200

-1.5

-1

-0.5

0

0.5

1

1.5

Time samples

Am

plitu

de


Figure 8.24a-3 Response with Manchester code at 3.2 kbps.

With the higher data, there is significant distortion of the signal due to the fact that the bandwidth of the signal exceeds that of the channel. (b) To included match filtering and plot the eye diagram, add the lines

DetectedSig = filter(pulse,1,RxSig); ploteye(DetectedSig,Ns);

after the signal has been passed through the channel model. The eye diagram with the 1.6 kbps Manchester code after matched filtering is shown below. There are two eyes present in the diagram, one at 0.6 and 1.6 symbol offsets. The eye is open to almost the full amplitude of the signal indicating that the signaling format is quite robust in the presence of noise. The eye, however, is quite narrow indicating it is not tolerant of large timing errors.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-20

-15

-10

-5

0

5

10

15

20

Symbol Periods

Figure 8.24b-1 Eye diagram with Manchester code at 1.6 kbps.

The eye diagram with the 3.2 kbps Manchester code after matched filtering is shown below. There are also two eyes present in the diagram, one at 0.8 and 1.8 symbol offsets. The eye is open, but the opening is significantly less than with the slower transmission rate, indicating a greater susceptibility to noise.


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-10

-8

-6

-4

-2

0

2

4

6

8

10

Symbol Periods

Figure 8.24b-2 Eye diagram with Manchester code at 3.2 kbps.

(c) Implement 4-ary signaling by changing

data = sign(randn(1,Nb)); to

data = 2*floor(4*rand(1,Nb)) - 3; which produces random symbols ±1, and ±3. The eye diagram with a 1.6 kHz symbol rate is shown below. The diagram shows three open eyes clearly separating the four levels and only a small amount of intersymbol interference.


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-60

-40

-20

0

20

40

60

Symbol Periods

Figure 8.24c-1 Eye diagram with Manchester code at 1.6 kbps and 4-ary signaling.

When the signaling rate is increased to 3.2 kHz, we obtain the eye diagram shown below

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-25

-20

-15

-10

-5

0

5

10

15

20

25

Symbol Periods

Figure 8.24c-2 Eye diagram with Manchester code at 3.2 kbps and 4-ary signaling.

In this case, there is no clear eye opening and transmission is unreliable, even in the absence of noise.


Problem 8.25 The eye diagram for the case of τ = T/2 and α=1.0 is shown below

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

0.04

Symbol Periods

Figure 8.25a-1 Eye diagram with α=1.0 and τ = T/2.

The eye is open approximately one-half of the full signal amplitude. When the rolloff is reduced to 0.5, we obtain the following eye diagram

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

0.04

Symbol Periods

Figure 8.26a-2 Eye diagram with α=0.5 and τ = T/2.


The height of the eye opening is relatively unchanged with α = 0.5 but we find the eye width is slightly less indicating a greater sensitivity to timing errors. (b) The results with τ = T are shown below for α = 1.0 and 0.5 respectively. The larger value of τ implies that the channel has a narrower bandwidth. This is turn causes more ISI, which becomes evident with the reduced eye opening.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

Symbol Periods Figure 8.25b-1 Eye diagram with α=1.0 and τ = T.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

Symbol Periods Figure 8.25b-2 Eye diagram with α=0.5 and τ = T.


Problem 8.26 We create the following function to compute the equalizer,

%----------------------------------------- % Problem 8.26 - Compute equalizer %------------------------------------------- function w = equalizer(h,pulse,N, Ns); %-- Compute system impulse response --- c = conv(h,pulse); % combine tx pulse shape and channel response [peak,centre] = max(abs(c)); % locate peak of impulse response and define as centre centre = round(centre); %--- Compute "C" matrix for two cases of N=3 or 5 --- if (N==3) C = [c(centre) c(centre-Ns) c(centre-2*Ns)]; C = [C; ... c(centre+Ns) c(centre) c(centre-Ns)]; C = [C; ... c(centre+2*Ns) c(centre+Ns) c(centre)]; b= [ 0 1 0]'; elseif (N==5) % note coefficient is zero if index is 0 or negative C = [c(centre) c(centre-Ns) c(centre-2*Ns) 0 0]; C = [C; ... c(centre+Ns) c(centre) c(centre-Ns) c(centre-2*Ns) 0]; C = [C; ... c(centre+2*Ns) c(centre+Ns) c(centre) c(centre-Ns) c(centre-2*Ns)]; C = [C; ... c(centre+3*Ns) c(centre+2*Ns) c(centre+Ns) c(centre) c(centre-Ns)]; C = [C; ... c(centre+4*Ns) c(centre+3*Ns) c(centre+2*Ns) c(centre+Ns) c(centre)]; b= [ 0 0 1 0 0]'; else ['N not supported'] end %--- Compute equalizer --- w = inv(C)*b; w = w/max(abs(w)); % Normalize equalizer return

To apply an equalizer to the signal of Problem 8.25, we modify the script to the following.

%------------------------------------------------------ % Prob 8.26 Equalized RC pulse shaping %------------------------------------------------------ T = 1; % symbol period Rs = 1/T; % symbol rate Ns = 16; % number of samples per symbol Fs = Rs*Ns; % sample rate (kHz) Nb = 1000; % number of bits to simulate alpha = 0.5; % rolloff of raised cosine N = 3; % number of equalizer taps %--- Discrete model of channel --- t = [0 : 1/Fs : 5*T]; h = exp(-t / (T)) /Fs; % impulse response scaled for sample rate pulse = firrcos(5*Ns, Rs/2, Rs*alpha, Fs); % 100% raised cosine filter %--- compute equalizer ---


w = equalizer(h,pulse,N, Ns); w = [w(:) zeros(N,Fs-1)]'; % upsample to Fs, so we can generate eye diagram w = w(:); %--- Pulse shape the data ----- data = sign(randn(1,Nb)); % random binary data Udata = [1; zeros(Fs-1,1)] * data; % upsample data Udata = Udata(:); % " Sig = filter(pulse,1,Udata); % pulse shape data Sig = Sig((length(pulse)-1)/2:end); % remove filter delay %--- Pass signal through the channel ---- RxSig = filter(h,1,Sig); %--- Equalize signal ---- EqSig = filter(w,1, RxSig); %--- Plot results ------------------------ ploteye(EqSig(4*Fs:end), Fs); % ignore initial transient xlabel('Symbol Periods')

Then with a N=3 tap equalizer, we obtain the following eye diagram. This is a huge improvement over the eye diagram without equalization that we observed in Problem 8.25b. The eye opening increases from 10% of the peak amplitude to approximately 80% of the peak amplitude at the sample instant.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-0.06

-0.04

-0.02

0

0.02

0.04

0.06

Symbol Periods

Figure 8.26a Eye diagram with α=0.5, τ = T, and N=3 equalizer.

When the number equalizer taps is increased to N=5, the further improvement for this channel is small as seen by the following eye diagram.


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-0.06

-0.04

-0.02

0

0.02

0.04

0.06

Symbol Periods

Figure 8.26b Eye diagram with α=0.5, τ = T, and N=5 equalizer.


Chapter 9

Problem 9.1 The three waveforms are shown below for the sequence 0011011001. (b) is ASK, (c) is PSK; and (d) is FSK.

Problem 9.2 The bandpass signal is given by ( )( ) ( ) cos 2 cs t g t f tπ= The corresponding amplitude spectrum, using the multiplication theorem for Fourier transforms, is given by

[ ]( ) ( )* ( ) ( )

( ) ( )c c

c c

S f G f f f f fG f f G f f

δ δ= − + +

= − + +

For a triangular spectrum G(f), the corresponding sketch is shown below. Problem 9.3 (a) From Example 5.12, we can compute the bandpass spectrum directly from the spectrum of the baseband equivalent representation. The baseband representation of the signal is

( ) ( ) ( )I Qg t g t jg t= + The autocorrelation of this complex random process is ( ) ( ) ( ) ( ) ( )g I Q IQ QIR R R jR jRτ τ τ τ τ= + − + The corresponding baseband power spectrum is then ( ) ( ) ( ) ( ) ( )I Q QI IQS f G f G f j G f G f⎡ ⎤= + + −⎣ ⎦ If the two random processes are independent and zero mean, then the cross-correlations are zero (and so are the corresponding cross-spectra). Then, the baseband spectrum is given by ( ) ( ) ( )I QG f G f G f= + The corresponding bandpass spectrum is then


[ ]14( ) ( ) ( )c cS f G f f G f f= − + +

(b) If gI(t) and gQ(t) are independent NRZ line codes then the corresponding baseband power spectra are 2( ) ( ) sinc( )I Q b bG f G f A T fT= = So bandpass spectrum is

( ) ( )2 214( ) sinc ( ) sinc ( )b c b b c bS f A T f f T A T f f T⎡ ⎤= − + +⎣ ⎦

And the spectrum looks like the following

Figure 9.3b. Bandpass spectrum with baseband NRZ line codes.

(c) If gI(t) = -gQ(t) then the signals are not independent, and the cross-spectral densities are not zero. However in this case ( ) ( )IQ QIR Rτ τ= and we obtain the same result, as if they were independent. (d) In this case where pulse shaping has a raised cosine spectral shape, the bandpass signal has the spectrum shown in the following.


Figure 9.3d. Bandpass spectrum with baseband pulse having a raised cosine spectral shape.

Problem 9.4





Problem 9.5


Problem 9.6

**The problem here is solved as “erfc” here and in the old edition, but listed in the textbook question as “Q(x)”.



Problem 9.7


Problem 9.8


Problem 9.9


Problem 9.10


Problem 9.11

Problem 9.12






Problem 9.13 Problem 9.13 A block diagram of a simple non-coherent detector consists of an energy integrator as shown in Figure 9.13(a). Note that the detector must be sampled and then cleared at the end of each bit interval

s(t)ξ Compare

toThreshold

2

0

( )bT

s t dt∫

EnergyDetector

Figure 9.13a Non-coherent detector.

A block diagram of a more complex coherent detector for ASK consists of a coherent down conversion to baseband followed by an integrate and dump detector. As with the energy detector of part (a), the integrate and dump detector must be sampled and then cleared at the end of each bit interval. The additional complexity with the coherent detector arises in the need for a carrier recovery circuit, and we are compensated for this additional complexity with better performance.

0

( )bT

s t dt∫ ξ Compareto

Threshold

CarrierRecovery

TimingRecovery

Integrate and DumpDetector

Figure 9.13b Coherent detector.


Problem 9.14

Problem 9.15



Problem 9.16

Problem 9.17


Problem 9.18

Problem 9.19 The important point to note here, in comparison to the plotted results, is that the error performance of the coherent QPSK is slightly degraded with respect to that of coherent PSK and coherent MSK. Otherwise, the observations made in Section 9.5 still hold here.


Problem 9.20


Problem 9.21 (a) The spectrum for the NRZ signal is given by (see Problem 8.3)

( )2 2( ) sincb bS f A T fT= This is the baseband equivalent spectrum of binary PSK. (b) The analytical expression for the MSK pulse shape is

cos

( ) 20

bb

t t Ts t T

otherwise

π⎧ ⎛ ⎞<⎪ ⎜ ⎟= ⎨ ⎝ ⎠

⎪⎩

The amplitude spectrum of this pulse shape is obtained by taking its Fourier transform.

[ ]

( )

( ) cos exp 22

cos cos 22

b

b

b

b

T

Tb

T

Tb

tH f j ft dtT

t ft dtT

π π

π π

−

−

⎛ ⎞= −⎜ ⎟

⎝ ⎠⎛ ⎞

= ⎜ ⎟⎝ ⎠

∫

∫

where the second line follows from the odd symmetry of sin(x). Using the identity cos A cos B = ½ [cos(A-B)+cos(A+B)] we obtain


( ) 1 cos 2 cos 22 2 2

sin 2 sin 22 21

2 2 22 2

2sin 2 sin 21 2 22 2

2 2

b

b

b

b

T

Tb b

T

b b

b bT

b b

b

H f f t f t dtT T

f t f tT T

f fT T

fT fT

fT

π ππ π

π ππ π

π ππ π

π ππ π

π ππ

−

−

⎧ ⎫⎛ ⎞ ⎛ ⎞⎡ ⎤ ⎡ ⎤⎪ ⎪= − + +⎜ ⎟ ⎜ ⎟⎨ ⎬⎢ ⎥ ⎢ ⎥⎜ ⎟ ⎜ ⎟⎣ ⎦ ⎣ ⎦⎪ ⎪⎝ ⎠ ⎝ ⎠⎩ ⎭

⎧ ⎫⎛ ⎞ ⎛ ⎞− +⎪ ⎪⎜ ⎟ ⎜ ⎟

⎪ ⎪⎝ ⎠ ⎝ ⎠= +⎨ ⎬⎪ ⎪− +⎪ ⎪⎩ ⎭

⎛ ⎞ ⎛ ⎞− +⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠= +

−

∫

2b

fT

π

⎧ ⎫⎪ ⎪⎪ ⎪⎨ ⎬⎪ ⎪+⎪ ⎪⎩ ⎭

Using the identity ( )sin cos2

x xπ⎛ ⎞± =⎜ ⎟⎝ ⎠

, the amplitude spectrum of the MSK pulse shape

simplifies to

( ) ( )

( )( )

2

22

cos 2 cos 2( )

2 22 2

cos 241 4

b b

b b

bb

b

fT fTH f

f fT T

fTTfT

π ππ ππ π

ππ

= +− +

=−

As described in Problem 9.3, a random sequence using this pulse shape will have a spectrum 2( ) ( )S f H f= .Comparing this to the sinc function which corresponds to the spectrum of a rectangular pulse, as shown in part (a), we find that the tails of the power spectrum decrease much faster, according to f4, with the MSK pulse shape.


Problem 9.22



Problem 9.23 (a) Let xI0 and xQ0 denote the in-phase and quadrature components of the matched filter output in the lower path of Figure 9.12 when a “1” is transmitted. Then the output of the enveloped detector is given by 2 2

0 0 0I Ql x x= + (1) Now the channel noise is w(t) is both white with power spectral density N0/2 and Gaussian with zero mean. Correspondingly, we find that the random variables XI0 and XQ0 (represented by samples xI0 and xQ0) are both Gaussian-distributed with zero mean and variance N0/2, given the phase θ. Hence we may write

0

20

000

1( ) expI

IX I

xf xNNπ

⎛ ⎞= −⎜ ⎟

⎝ ⎠ (2)

and

0

20

000

1( ) expQ

QX Q

xf x

NNπ

⎛ ⎞= −⎜ ⎟⎜ ⎟

⎝ ⎠ (3)

Note that XI0 and XQ0 are independent Gaussian random variables, and so we may express there joint probability density function by

( )0 0

2 20 0

, 0 00 0

1, expI Q

I QX X I Q

x xf x x

N Nπ⎛ ⎞+

= −⎜ ⎟⎜ ⎟⎝ ⎠

(4)

Recall the rectangular-to-polar conversion 0 0 cosIx l θ= (5) 0 0 sinQx l θ= (6) In a limiting sense, we may equate the two areas of the different co-ordinate systems 0 0 0 0I Qdx dx l dl dθ= (7) where l0 and θ are the envelope and phase of the observed process. Then substituting (5) and (6) into (4), we find that the probability of the random variables L0 and Θ lying in the area defined by (7) is

2

0 00

0 0

expl l dl dN N

θπ

⎡ ⎤−⎢ ⎥⎣ ⎦

(8)

and the joint probability density function of L0 and Θis given by

( )0

20 0

, 00 0

, expLl lf lN N

θπΘ

⎡ ⎤= −⎢ ⎥

⎣ ⎦ (9)

This probability function is independent of the angle θ, and consequently L0 and Θ are statistically independent, i.e., ( ) ( ) ( )

0 0, 0 ,L L of l f l fθ θΘ Θ= . In particular, the phase is uniformly distributed inside the range 0 to 2π as shown by


1 0 2

( ) 20

felsewhere

θ πθ πΘ

⎧ ≤ ≤⎪= ⎨⎪⎩

(10)

This leaves the probability density function of the random variable L0 as

( )0

20 0

00 0 0

2 exp 0

0L

l l lf l N N

elsewhere

⎧ ⎡ ⎤− ≥⎪ ⎢ ⎥= ⎨ ⎣ ⎦

⎪⎩

(11)

which is the Rayleigh probability density function. (b) The output of the upper envelope detector of Figure 9.12, when a “1” is sent, is the equivalent of a sinusoid plus noise. A sample function of the sinusoidal wave plus noise is then expressed by ( )( ) cos 2 ( )c cx t A f t n tπ= + (12) Representing the narrowband noise n(t) in terms of its in-phase and quadrature components, we may write ( ) ( )'( ) ( ) cos 2 ( )sin 2I c Q cx t n t f t n t f tπ π= + (13) where ' ( ) ( )I In t A n t= + (14) We assume that n(t) is Gaussian with zero mean and variance N0/2. Accordingly, we may state the following:

(i) Both ' ( )In t and ( )Qn t are Gaussian and statistically independent.

(ii) The mean of ' ( )In t is A and that of ( )Qn t is zero.

(iii) The variance of both ' ( )In t and ( )Qn t is N0/2.

We may therefore express the joint probability density function of the random variables 'IN and QN corresponding to ' ( )In t and ( )Qn t as follows;

( )

'

2' 2'

,0 0

1( , ) expI Q

I QI QN N

n A nf n n

N Nπ

⎡ ⎤− +⎢ ⎥= −⎢ ⎥⎣ ⎦

(15)

Let l1 denote the envelope of x(t) and θ denote its phase. From the complex baseband equivalent of Eq. (13) we find that

( )2' 21 I Ql n n= + (16)

and

1'tan Q

I

nn

θ − ⎡ ⎤= ⎢ ⎥

⎣ ⎦ (17)


Following a procedure similar to that described in the derivation of the Rayleigh distribution, we find that the joint probability density function of the random variables L1 and Θ, corresponding to l1 and θ for some fixed time t, is given by

1

2 21 1 1

,0 0

2 cos( , ) expLl l A Alf rN N

θθπΘ

⎡ ⎤+ −= −⎢ ⎥

⎣ ⎦ (18)

We see that in this case, however, we cannot express the joint probability density function , ( , )Rf r θΘ as a product ( ) ( )Rf r f θΘ . This is because we now have a term involving the values of both random variable multiplied together as rcos θ. Hence, L1 and Θ, are dependent. We are interested, in particular, in the probability density function of L1. To determine this probability density function, we integrate Eq. (18) over all possible values of θ, obtaining the marginal density

1 1

2

1 , 10

2 2 21 1 10

0 0 0

( ) ( , )

2 21exp exp cos2

RLf l f l d

l l A Al dN N N

π

π

θ θ

θ θπ

Θ=

⎡ ⎤ ⎛ ⎞+= − ⎜ ⎟⎢ ⎥

⎣ ⎦ ⎝ ⎠

∫

∫ (19)

The integral on the right-hand side of Equation ( ) can be identified in terms of the defining integral for the modified Bessel function of the first kind of zero order (see Appendix), that is

( )2

0 0

1( ) exp cos2

I x x dπ

θ θπ

= ∫ (20)

Thus, letting 21x Al σ= , we may rewrite Eq. (19) in the compact form

1

2 21 1 1

1 00 0 0

2 2( ) expLl l A Alf l I

N N N⎡ ⎤ ⎛ ⎞+

= − ⎜ ⎟⎢ ⎥⎣ ⎦ ⎝ ⎠

for 1 0l ≥ (21)

This is the Rician distribution. (c) The probability of an error is the probability that L0 > L1 when a “1” is transmitted and so

[ ] [ ]10 1 0 1 1 1 10

|1 | ( )LL L sent L l l p l dl∞

> = >∫P P (22)

where


[ ]1

1

20 0

0 1 1 00 0

20

0

21

0

2| exp

exp

exp

l

l

l lL l l dlN N

lN

lN

∞

∞

⎡ ⎤> = −⎢ ⎥

⎣ ⎦

⎡ ⎤= − −⎢ ⎥

⎣ ⎦

⎡ ⎤= −⎢ ⎥

⎣ ⎦

∫P

(23)

Substituting this result in Eq. (22), we obtain

[ ]

2 2 21 1 1 1

0 1 0 100 0 0 0

2 21 1 1

0 100 0 0

2 2|1 exp exp

2 2 2exp

l l l A AlL L sent I dlN N N N

l l A AlI dlN N N

∞

∞

⎡ ⎤ ⎡ ⎤ ⎡ ⎤+> = − −⎢ ⎥ ⎢ ⎥ ⎢ ⎥

⎣ ⎦ ⎣ ⎦ ⎣ ⎦⎡ ⎤ ⎡ ⎤+

= −⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦

∫

∫

P(24)

Define a new random variable

1

0

2lvN

= (25)

Then this last equation becomes

[ ]2 2

120 1 00

0 0

2 2201

2 000 0

|1 exp2

/exp exp2 2

v A AvL L sent v I dvN N

v A NA Avv I dvN N

∞

∞

⎡ ⎤⎡ ⎤⎛ ⎞> = − + ⎢ ⎥⎢ ⎥⎜ ⎟

⎢ ⎥ ⎢ ⎥⎝ ⎠⎣ ⎦ ⎣ ⎦⎡ ⎤⎡ ⎤ ⎡ ⎤+

= − − ⎢ ⎥⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎢ ⎥⎣ ⎦ ⎣ ⎦

∫

∫

P

(26)

The integrand in the last expression is the normalized Rician density function and when integrated over its range, has a value of unity. Thus,

[ ]2

120 1

0

|1 exp2AP L L sentN

⎡ ⎤> = −⎢ ⎥

⎣ ⎦ (27)

and with 2cAA = we obtain the answer given in the textbook.

Soln Manual 5e

Documents

matlab script

matlab script

rms phase

2t 2t

fm modulated

fm demodulated

tb ta

ja cos